CN101414551B - Reduction of etch-rate drift in hdp processes - Google Patents

Abstract

The present invention relates to the reduction of etch rate excursions in high density plasma processes. The present invention provides for aging a processing chamber by providing a gas flow of an aging precursor to the processing chamber. A high density plasma is formed from the aged precursor by applying a source power of at least 7500W, wherein greater than 70% of the source power is distributed at the top of the processing chamber. Utilizing said high density plasma deposition at one point having at least

The thickness of the aging layer. Each of the plurality of substrates is delivered sequentially into a processing chamber to perform a process including etching on each of the plurality of substrates. The processing chamber is cleaned between sequential transfers of substrates.

Description

Reduction of Etch Rate Shift in High Density Plasma Processes

Cross references to related applications

This application is a non-provisional patent application of U.S. Patent Application No. 60/970,884 filed on September 7, 2007 by Anchuan Wang et al., entitled "INTEGRATED PROCESSMODULATION", and claims priority on the filing date of this U.S. patent application right. Its entire contents are hereby incorporated by reference.

This application is also related to the concurrently assigned and commonly assigned U.S. Patent Application No. 12/204,523 by Anchuan Wang et al., entitled "IMPURITY CONTROL IN HDP-CVDDEP/ETCH/DEP PROCESS", and related to Anchuan Wang et al. in 2007 "GAPFILL EXTENSION OF HDP-CVD INTEGRATED PROCESS MODULATION SIO ₂ PROCESS", submitted on June 4, the entire content of each is hereby incorporated by reference.

technical field

The present invention relates to high density plasma technology.

Background technique

One of the persistent challenges faced in the development of semiconductor technology is the need to increase the density and interconnection of circuit elements on a substrate without introducing spurious interactions between circuit elements. Undesirable interactions are typically prevented by providing gaps and trenches filled with an electrically insulating material to physically and electrically isolate the elements. As circuit density increases, however, the width of these gaps decreases, which increases the aperture ratio of the gaps and makes it more difficult to fill the gaps without leaving voids. Formation of voids when the gap is not completely filled is undesirable because they can adversely affect the operation of the completed device, such as by trapping impurities within the insulating material.

A common technique employed in this gapfill application is the chemical vapor deposition ("CVD") technique. Conventional thermal CVD processes provide reactive gases to the substrate surface where thermally induced chemical reactions occur to form the desired film. Plasma-enhanced CVD ("PECVD") techniques generate a plasma by applying radio frequency ("RF") energy to a reaction zone near a substrate surface to induce excitation and/or dissociation of reactive gases. The high reactivity of the species in the plasma reduces the energy required for chemical reactions to occur, thus lowering the temperature required for the CVD process compared to conventional thermal CVD processes. These advantages can be further exploited using High Density Plasma ("HDP") CVD techniques, where dense plasmas are formed at low vacuum pressures so that the plasma species are even more reactive. While each of these techniques falls broadly under the general term "CVD techniques," each of them has characteristics that make them more or less suitable for certain applications.

HDP-CVD systems form plasmas that are at least about 2 orders of magnitude denser than standard, capacitively coupled plasma CVD systems. Examples of HDP-CVD systems include inductively coupled plasma systems and electron cyclotron resonance (ECR) plasma systems, among others. HDP-CVD systems generally operate at lower pressures than low density plasma systems. The low chamber pressures used in HDP-CVD systems provide active species with long mean free paths and reduced angular distributions. These factors, together with the plasma density, help a considerable fraction of the plasma to reach even the deepest parts of the closely spaced gaps and provide a Films with Improved Gap Fill Capability.

Another factor that allows films deposited using HDP-CVD techniques to have improved gap-fill characteristics is the simultaneous deposition of films by high-density plasma sputtering. The sputtering component of the HDP deposition process slows deposition on certain features such as the corners of raised surfaces, resulting in an increase in the gap-fill capability of the HDP deposited film. Some HDP-CVDF systems introduce argon or a similar heavy inert gas to further enhance the sputtering effect. These HDP-CVD systems typically utilize electrodes within a substrate support pedestal that are capable of creating an electric field to bias the plasma toward the substrate. An electric field can be applied throughout the HDP deposition process to further facilitate sputtering and provide better gap-fill characteristics for known films.

It was initially thought that due to the nature of simultaneous deposition/sputtering, HDP-CVD could fill gaps or trenches created in almost any application. However, semiconductor manufacturers have discovered that there is a practical limit to the aperture ratio of the gaps that the HDP-CVD process can fill. For example, one HDP-CVD process commonly used to deposit silicon oxide gap-fill films forms a plasma from process gases including silane _SiH4 , molecular oxygen _O2, and argon Ar. It has been reported that when the process is used to fill certain narrow-width high-aspect ratio gaps, sputtering caused by argon in the process gas can hinder the gap-fill effect. In particular, it has been reported that species sputtered by in-process argon redeposit at a faster rate on the upper portion of the sidewall of the gap being filled than on the lower portion of the sidewall of the gap being filled. Conversely, if the upper region of regrowth engages before the gap is completely filled, it can cause a void to form in the gap.

Figure 1 presents schematic cross-sectional views of silicon oxide films at different stages of deposition to illustrate possible gapfill limitations associated with some CVD processes. The gap fill problem is shown in slightly enlarged form to better illustrate the problem. The top of FIG. 1 shows the initial structure 104 with a gap 120 defined by two adjacent features 124 and 128 having a horizontal surface 122 , and a horizontal surface at the bottom of the gap labeled 132 . As shown in structure 108, i.e., the second portion from the top of the figure, a conventional HDP-CVD silicon oxide deposition process results in a direct deposition. However, indirect deposition (referred to as “re-deposition”) on the sidewall 140 of the gap 120 is also caused due to recombination of material sputtered from the silicon oxide film as the silicon oxide film grows. In certain small width, high aspect ratio applications, the continued growth of the silicon oxide film results in formations 136 on the upper portion of the sidewall 140 that grow toward each other at a growth rate that exceeds the lateral growth of the film on the lower portion of the sidewall. This trend is shown in structures 108 and 112, and the end result in structure 116 is the formation of voids 144 within the film. The probability of voiding is very directly related to the rate and character of redeposition.

Accordingly, there remains a general need in the art for improved gap filling techniques.

Contents of the invention

Embodiments of the invention provide methods of depositing films on a plurality of substrates. In a first set of embodiments, the processing chamber is seasoned by providing a flow of a season precursor to the processing chamber. A high density plasma is formed from the aged precursor by applying a source power of at least 7500W, with greater than 70% of the source power distributed at the top of the processing chamber. Utilizes high density plasma deposition with at least 5000 at one point The thickness of the aging layer. Each of the plurality of substrates is transferred sequentially into the processing chamber to perform a process including etching on each of the plurality of substrates. The processing chamber is cleaned between sequential transfers of each of the plurality of substrates.

In various embodiments, at least 5000 The thickness can include at least 7500 thickness or may include at least 10,000 thickness of. The aging precursor gas flow may be provided as a silicon-containing gas flow such as _SiH4 and an oxygen-containing gas flow such as _O2 . The flow rate of the oxygen-containing gas may be less than the flow rate of the silicon-containing gas, or may be 0.8 times or less than the flow rate of the silicon-containing gas. In some instances, a gas flow that is non-reactive with the silicon-containing gas and the oxygen-containing gas is additionally provided, sometimes at a flow rate of less than 200 sccm.

In a second set of embodiments, the processing chamber is aged, and each of the plurality of substrates is sequentially transferred into the processing chamber to perform on each of the plurality of substrates, including etching craft. The processing chamber is cleaned between sequential transfers of each of the plurality of substrates by performing a partial cleaning of the processing chamber, then heating the processing chamber, and then completing cleaning of the processing chamber.

Performing a partial cleaning of the processing chamber and completing the cleaning of the processing chamber each include flowing a halogen precursor into the processing chamber and forming a high density plasma from the halogen precursor. An example of a suitable halogen precursor is _F2 . A spot cleaning exceeding 75% of the endpoint of the clean can be performed. The processing chamber may be heated by flowing heating gas into the processing chamber and forming high-density plasma from the heating gas. Examples of heating gas include O ₂ , Ar, He, and the like. A high-density plasma can be formed by applying source power that is approximately equally divided between the top and side sources.

A further understanding of the nature and advantages of the invention may be realized by referring to the remaining portions of the specification and drawings.

Description of drawings

Figure 1 presents a schematic cross-sectional view illustrating the formation of voids during a prior art gapfill process;

2 is a simplified cross-sectional view of a partially completed integrated circuit including a plurality of shallow trench isolation structures;

Figures 3A and 3B are schematic diagrams illustrating the distribution of gap-filling features in open regions and densely packed regions in structures;

Figure 4A is a flowchart outlining the process for depositing a film on a substrate in an embodiment of the invention;

FIG. 4B is a flow diagram outlining a specific deposition process that may be used with the method of FIG. 4A in which deposition steps and etch steps alternate;

4C is a flowchart outlining a method of aging a processing chamber that may be employed in certain embodiments as part of the method of FIG. 4A;

Figure 4D is a flowchart outlining a method of cleaning a processing chamber that may be employed in certain embodiments as part of the method of Figure 4A;

Figure 5A is a simplified diagram of one embodiment of a high density plasma chemical vapor deposition system that can implement the methods of the present invention;

5B is a simplified cross-sectional view of a gas ring that may be used with the exemplary processing system of FIG. 5A.

Detailed ways

Embodiments of the invention relate to methods of depositing films on substrates incorporating an etching step. In a specific exemplary application of the present invention, a method of depositing a silicon oxide layer using a high density plasma CVD process to fill gaps in a substrate surface is provided. The silicon oxide film deposited according to the technique of the present invention has excellent gap-fill capability and is capable of filling, for example, gaps occurring in shallow trench isolation ("STI") structures. The films deposited by the method of the invention are thus suitable for use in the manufacture of a wide variety of integrated circuits, including those having feature sizes on the order of less than 45 nm.

One discovery of the inventors, as part of their research on this deposition method, is that during the etch portion of the method there is a systematic tendency for the etch rate to decrease as more substrate is processed. "Etch-rate drift" occurs even when the process conditions for each substrate are substantially the same.

Figure 2 illustrates the types of structures that may be populated according to embodiments of the present invention, and Figure 2 presents a simplified cross-sectional view of a partially completed integrated circuit 200 . The integrated circuit is formed on a substrate 204 comprising a plurality of STI structures, where each STI structure is typically formed by forming a thin pad oxide layer 220 on the surface of the substrate 204 and then forming a nitride oxide layer on the pad oxide layer 220. Silicon layer 216 is formed. The nitride and oxide layers are then patterned using standard photolithographic techniques, and trenches 224 are etched through the nitride/oxide stack in substrate 204 . FIG. 2 shows an integrated circuit that may include a relatively densely packed region 208 of transistors or other active devices, and may include a relatively isolated open region 212 . The active devices in the open area 212 may be separated from each other by more than an order of magnitude more separation than in the densely packed area 208, but an "open area" as used herein is considered to be a space in which there is a gap. A region whose width is at least five times the width of the gap in the "dense region".

Embodiments of the present invention provide methods of filling trenches 224 with an electrically insulating material, such as silicon dioxide, using a deposition process with better gapfill properties. In some examples, prior to the gapfill process, an initial liner layer is deposited on the substrate as an in-situ steam generation ("ISSG") or other thermal oxide layer, or may be a silicon nitride layer. One advantage of depositing this liner before filling trenches 224 is to provide proper fillet, which can help avoid the described effects of early gate breakdown in the formed transistor.

As used herein, a high-density plasma process is a plasma CVD process that involves simultaneous deposition and sputtering of components and applies a plasma with an ion density on the order of 10 ¹¹ ions/cm ³ or greater. The relative levels of the combined deposition and sputtering characteristics of a high-density plasma can depend on factors such as the flow rate used to deliver the gas mixture, the source power level applied to maintain the plasma, the bias power applied to the substrate, etc. . The combination of these factors can be conveniently quantified by the "deposition/sputter ratio", sometimes expressed as D/S, to define the characteristics of the process:

The deposition/sputtering ratio increases with increasing deposition and decreases with increasing sputtering. As used in the definition of D/S, "net deposition rate" refers to the deposition rate measured when deposition and sputtering occur simultaneously. The "blanket sputtering rate" is the sputtering rate measured when the process recipe is executed without deposition gas; the pressure in the process chamber is adjusted to the pressure during deposition, and the sputtering rate is Measured on patterned thermal oxide.

Other equivalent measurements can be used to quantify the relative deposition and sputtering contributions of the HDP process, as known to those skilled in the art. A commonly used selectable ratio is the "etch/deposition ratio",

This "etch/deposition ratio" increases with increasing sputtering and decreases with increasing deposition. As used in the definition of E/D, "net deposition rate" again refers to the deposition rate measured when deposition and sputtering occur simultaneously. However, "source-only deposition rate" refers to the measured deposition rate when the process recipe is executed without sputtering. Embodiments of the invention are described herein in terms of D/S ratio. While D/S and E/D are not exact inverses, they are inversely related, and the conversion between them will be understood by those skilled in the art.

The desired D/S for the known steps in the HDP-CVD process is typically obtained by including a precursor gas flow that may also function as a sputtering agent, and in some instances a flowing gas flow that acts as a sputtering agent ratio. The elements contained by the precursor gases react to form a film with the desired composition. For example, to deposit a silicon oxide film, a precursor gas may include a silicon-containing gas such as silane SiH ₄ , and an oxidizing gas reactant such as molecular oxygen O ₂ . Dopants can be added to the film by including a precursor gas with the desired dopant, such as by including a gas flow of SiF to fluorinate the film, a gas flow of pH ₃ to _phosphate the film, _a gas flow of _B2H6 to boride the membrane, gas flow including _N2 to nitride the membrane, etc. The flowing gas can be provided with a flow of _H2 or an inert gas flow including a He flow, or an even heavier inert gas flow such as Ne, Ar or Xe. The level of sputtering provided by different flowing gases is directly related to their atomic mass (or molecular mass in the case of _H2 ), with _H2 producing even less sputtering than He. Embodiments of the invention generally provide a flowing gas stream having an average molecular mass of less than 5 amu. This can be achieved by utilizing a single low mass gas, such as a flow of substantially pure _H2 or a flow of substantially pure He. Alternatively, gas flow can sometimes be provided by multiple gases, such as by providing both _H2 and He gas flows, which are mixed in the HDP-CVD processing chamber. Alternatively, sometimes the gases can be premixed so that the _H2 /He gas stream is provided to the processing chamber in a mixed state. It is also possible to provide a separate stream of the higher mass gas, or to include the higher mass gas in the premix, with the relative flow rates and/or concentrations of the premix selected to maintain an average molecular mass of less than 5 amu.

In high aperture ratio structures, in general, relatively high flow rates of low mass flowing gases have been found to improve gap fill capability compared to more traditional use of flowing gases such as Ar. This is believed to be a result of the reduction in redeposition achieved by utilizing He or _H2 as the flowing gas. But even with such low quality flowing gases, there is a risk of comer clipping during deposition. This effect can be understood with reference to FIGS. 3A and 3B , which illustrate the effect of the sputtering component of the HDP process for gaps in densely packed regions and for gaps in open regions, respectively.

Specifically, gap 304 in FIG. 3A is a high aspect ratio gap having material deposited using a HDP-CVD process that forms cust structures 308 on horizontal surfaces. Redeposition occurs when material 312 is sputtered from tip 308 in response to collisions of plasma ions along path 316 . The sputtered material 312 follows a path 320 to a sidewall 324 on the opposite side of the gap 304 . The effect is symmetrical so that when material is sputtered away from the left side of the gap to the right, material is also sputtered away from the right side of the gap to the left. Redeposition of material protects against excessive sputtering that can cause corner beading.

In an open area, such as open area 330 as shown in Figure 3B, this symmetry does not exist. In this example, deposition results in formation similar to tip 308', but when material 312' is sputtered along line 320' in response to collisions of plasma ions along line 316', the opposite sides of the gap are too far away to occur with Protective redeposition. The corners of the structure in Figure 3B are subject to sputtering of the same material as the corners of the structure in Figure 3A, but without the compensating effect of receiving material sputtered from the opposite side of the gap. Therefore, there is an increased risk of crimping corners and damaging underlying structures.

The method of the invention is summarized in the flowchart of Figure 4A, which provides an overview of the method. These methods apply to the operation of several substrates in a common processing chamber, where multiple processes are performed on each substrate. At block 402, the method begins by aging the processing chamber, ie, coating the internal structure of the processing chamber with a material, one example of which includes _SiO2 . At block 404 , the substrate is transferred into a processing chamber so that at block 406 a process is performed on that substrate. The process includes significant etching even when the overall result of application of the process is a net deposition of material. At block 410 , the substrate is transferred out of the processing chamber, and at block 412 the processing chamber is cleaned.

A check is performed at block 414 to see if the entire substrate operation has been completed. A typical substrate run might involve five substrates, although a greater or lesser number of substrates may be used in different specific embodiments. If the substrate operation is not complete, the next substrate in the operation is transferred into the processing chamber at block 404 and the method is repeated for the next substrate. Once all substrate operations have been completed, the processing chamber is aged again at block 402 using the same process as at block 406 or using a different process in preparation for another substrate operation.

FIG. 4B gives details of those processes that may be employed at block 406 . In this example, the deposition is obtained on the substrate using a deposition/etching/deposition process, but more generally the method of the invention can be applied to other types of processes with a significant etching component. The substrate is typically a semiconductor wafer such as a 200mm or 300mm diameter.

At block 420, a precursor gas flow is provided to the chamber, the precursor gas flow including a silicon precursor flow, an oxygen precursor flow, and a flow gas flow. Table 1 provides exemplary flow rates for depositing undoped silicate glass ("USG") using gas streams of monosilane _SiH4 , molecular oxygen O2 _, and _H2 , but it should be understood that Other precursor gases including dopant sources and other flow gases may be used.

Table I: Exemplary flow rates for USG deposition

As shown, the flow rates of the precursor gases are similar for 200-mm and 300-mm diameter wafers, but the flow rates of the flowing gases are generally higher.

(埃/分钟)的相对低的沉积速率。发明人已经发现对于很小的特征尺寸，一般利用低沉积速率和高D/S比率的这种组合来改善间隙填充特征。In block 422, a high concentration plasma is formed from a gaseous flow by coupling energy into the chamber. A common technique used to generate high-concentration plasmas is inductively coupling radiofrequency energy. The D/S ratio is determined not only by the flow rate of the gas, but also by the power density coupled into the chamber, by the strength of the bias that can be applied to the substrate, by the temperature inside the chamber, by the pressure inside the chamber determined, and determined by other factors. For the deposition of the initial portion of the film in block 424, in some embodiments, the process parameters may be selected to provide a D/S ratio in excess of 20 while simultaneously providing a D/S ratio of 900-6000

(Angstroms/min) relatively low deposition rate. The inventors have found that for very small feature sizes, this combination of low deposition rate and high D/S ratio is generally utilized to improve gapfill characteristics.

After deposition is complete, at block 426, the deposition precursor flow is terminated, and then at block 428 it is checked whether the film has reached the desired thickness. Embodiments of the present invention include at least two deposition stages separated by an etch stage, frequently there may be 5-15 deposition stages or even more deposition stages depending on the specific characteristics of the gap being filled.

At block 430, the etch phase of the process begins by flowing a halogen precursor, typically including a fluorine precursor such as NF ₃ or a chlorofluorocarbon. At block 432, a high concentration plasma is formed from the halogen precursor using a high source power density. In some embodiments, the source power density is between about 80,000 and 140,000 W/m ² , which equates to a total source power of between about 6000 and 10,000 watts for a 300-mm diameter wafer, and for a 200-mm diameter wafer This equates to a total source power of between about 2500 and 4500 watts. The inventors have found that using high source power makes the deposition profile more symmetrical than using low source power. In some embodiments, the total source power is distributed among the top and side sources such that a major portion of the source power is provided by the side sources. For example, the side source power may be 1-5 times the top source power, and in a particular embodiment, the side source power is 3 times the top source power.

In block 434, the generated halogen plasma is used to etch back the deposited film. Although the specific amount of material that can be etched is relatively dependent on the specific structure of the substrate structure, typically the amount of material that can be etched in later etch cycles is greater than the amount of material etched in earlier etch cycles. This is a general consequence of the fact that the overall layout of the substrate changes due to the sequence of deposition and etching steps. The general tendency of this sequence of steps is that the layout becomes more etched in larger volumes during the etch phase of the cycle. At block 436, the flow of the halogen precursor is terminated, and at block 420, the process may return to the deposition phase by re-flowing the flow of silicon precursor, oxygen precursor and flowing gas.

之间，同时当每个周期使用更大的沉积量时整个工艺需要更少的周期。为了沉积相同量的材料，当每个周期沉积300

时，需要在每个周期沉积1000

时大约六倍的周期。It is generally expected that the same precursors will be used for the deposition of material during each deposition phase and the same precursors will be used for removal of material during the etch phase, but this is not necessary for the present invention. The amount of material deposited in each deposition stage is usually between 300 and 1000

between, while the overall process requires fewer cycles while using larger deposition volumes per cycle. To deposit the same amount of material, when 300

, it is necessary to deposit 1000 per cycle

about six times as long.

FIG. 4C illustrates an aging process that may be used at block 402 in some embodiments. At block 440, the aging process begins to establish aging conditions within the chamber. In some embodiments, the conditions include a chamber pressure between 25 and 65 millitorr (mtorr). A gas flow of an aging precursor is provided at block 442. In embodiments where the aging process includes a _SiO2 coating, the precursor may include a silicon-containing gas such as silane and an oxygen-containing gas. For example, the silicon precursor may include _SiH4 and the oxygen precursor may include _O2 . In some embodiments, the flow rate of the oxygen-containing gas is less than the flow rate of the silicon-containing gas, may be less than 0.9 times the flow rate of the silicon-containing gas, may be less than 0.8 times the flow rate of the silicon-containing gas, and may be less than 0.8 times the flow rate of the silicon-containing gas. The flow rate of the gas is 0.7 times or less, may be 0.6 times or less than the flow rate of the silicon-containing gas, or may be 0.5 times or less than the flow rate of the silicon-containing gas. For example, in one embodiment using _SiH4 and _O2 , the flow rate of _O2 is 300 seem and the flow rate of _SiH4 is 470 seem, a flow ratio equal to approximately 0.65. The gas flow provided at block 442 may also sometimes include non-reactive gases, such as embodiments using He, Ne, or Ar. The flow rate of the non-reactive gas is typically less than 200 seem to reduce sputtering effects, and may be 0 seem.

At block 444, a high density plasma is formed from the aged precursors by coupling energy into the processing chamber as described above. The energy is preferably coupled preferentially with the application of top source power, and in embodiments of the invention, the energy has more than 70% of the applied source power at the top of the chamber and more than 80% of the applied source power at the top of the chamber , more than 90% of the source power applied at the top of the chamber, or even 100% of the source power applied at the top of the chamber. The typical power applied is greater than 7500W, and in one embodiment, approximately 9000W of power applied overall at the top of the processing chamber is utilized.

的厚度，在一个点处具有至少6000

的厚度，在一个点处具有至少7500的厚度，在一个点处具有至少10,000

的厚度，或者在一个点处具有至少12,500的厚度。在时效处理层沉积后，在区块448，停止时效前驱物气流。At block 446, the high density plasma is used to deposit an aging treatment layer. While it is generally expected that the thickness of the aged layer may be non-uniform, in some embodiments it has at least 5000

thickness, at one point having at least 6000

thickness, at one point having at least 7500 thickness, at one point having at least 10,000

thickness, or at one point have at least 12,500 thickness of. After the aging layer is deposited, at block 448, the gas flow of the aging precursor is stopped.

FIG. 4D presents a flowchart of one method of cleaning the chamber at block 412 of FIG. 4A in accordance with an embodiment of the present invention. The cleaning method includes two-step cleaning, and a heating process is performed between the two cleaning steps.

Accordingly, at block 460, after the substrate has been processed, a halogen precursor, such as F2, is flowed into the processing chamber. At block 462, a high density plasma is formed from the halogen precursor, and at block 464, a local clean is performed using the high density plasma. Local cleaning of greater than 75% of the process endpoint may be performed in some embodiments.

After the first cleaning step is complete, the flow of the halogen precursor is stopped at block 466 . At block 468, this gas flow is replaced with a heated gas flow from which a heated plasma is formed at block 470. By way of example only, in various embodiments, the heating gas may include O2, Ar, and/or He, and an exemplary source power of 12,000 W is applied with equal top and side distribution for a period of 30 to 120 seconds. This intermediate heating acts to counteract the general cooling of the chamber that occurs during cleaning.

After stopping the heating gas flow at block 472, the halogen precursor gas flow may be reapplied at block 474 and the high density plasma again formed at block 475 for complete chamber cleaning.

Exemplary Substrate Processing System

The inventors have practiced embodiments of the invention with the ULTIMA ^™ system manufactured by APPLIED MATERIALS, INC. of Santa Clara, California, a general description of which is provided in commonly assigned In US Patent No. 6,170,428, the patent application was filed in 1996 by Fred C. Redeker, Farhad Moghadam, Hirogi Hanawa, Tetsuya Ishikawa, Dan Maydan, Shijian Li, Brian Lue, Robert Steger, Yaxin Wang, Manus Wong and Ashok Sinha It was submitted on the 15th, and its invention name is: "SYMMETRIC TUNABLE INDUCTIVELY COUPLED HDP-CVDREACTOR", here, its entire disclosure is incorporated as a reference. An overview of this system is provided below in connection with Figures 5A and 5B. Figure 5A schematically illustrates the structure of such an HDP-CVD system 510 in one embodiment. System 510 includes chamber 513 , vacuum system 570 , source plasma system 580A, bias plasma system 580B, gas delivery system 533 , and remote plasma cleaning system 550 .

The upper portion of the chamber 513 includes a dome 514 made of a ceramic dielectric material such as aluminum oxide or aluminum nitride. Dome 514 defines an upper boundary of plasma processing region 516 . The plasma processing region 516 is bounded at the bottom by the upper surface of the substrate 517 and the substrate support 518 .

A heating plate 523 and a cooling plate 524 are above and thermally connected to the dome 514 . The heating plate 523 and cooling plate 524 allow the dome temperature to be controlled within about ±10°C over a range of about 100°C to 200°C. This allows the dome temperature to be optimized for various processes. For example, it is desirable to keep the dome temperature higher for cleaning or etching processes than for deposition processes. Precise control of the dome temperature also reduces flaking or particle load within the chamber and improves adhesion between the deposited layer and the substrate.

The lower portion of the chamber 513 includes a body member 522 that connects the chamber to a vacuum system. The base portion 521 of the substrate support 518 is mounted on a body member 522 and forms a continuous inner surface with the body member 522 . Substrates are transferred into and out of chamber 513 by a robotic pallet (not shown) through insertion/removal openings (not shown) in the side walls of chamber 513 . The lift pins (not shown) are raised and then lowered under the control of a motor (also not shown) to move the substrate from the robot pallet at the upper mounting position 557 to the lower processing position 556 where the substrate is placed on the substrate receiving portion 519 of the substrate support 518 . The substrate receiving portion 519 includes an electrostatic chuck 520 that secures the substrate to the substrate support 518 during substrate processing. In a preferred embodiment, the substrate support 518 is made of an aluminum oxide or aluminum ceramic material.

The vacuum system 570 includes a throttle body 525 that houses a twin-blade throttle valve 526 and is attached to a gate valve 527 and a turbomolecular pump 528 . It should be noted that the throttle body 525 provides minimal obstruction to airflow and allows symmetrical pumping. Gate valve 527 can allow isolation of pump 528 from throttle body 525 and can control chamber pressure by limiting the exhaust capacity when throttle valve 526 is fully open. The arrangement of the throttle valve, gate valve and turbomolecular pump allows precise and stable control of the pressure of the chamber up to about 1 mTorr to about 2 mTorr.

Source plasma system 580A includes top coil 529 and side coil 530 mounted on dome 514 . A symmetrical ground shield (not shown) reduces electrical coupling between the coils. The top coil 529 is powered by a top source radio frequency (SRF) generator 531A, while the side coil 530 is powered by a side SRF generator 531B, and allows independent power levels and operating frequencies for each coil. This dual coil system allows control of the radiation ion concentration within the chamber 513, thereby improving plasma uniformity. The side coil 530 and top coil 529 are typically driven inductively, which does not require auxiliary electrodes. In a particular embodiment, top source RF generator 531A provides up to 2,500 Watts of RF power at nominal 2MH, while side source RF generator 531B provides up to 5,000 Watts of RF power at nominal 2MH. The operating frequencies of the top and side RF generators can be deviated from nominal operating steps (eg, to 1.7-1.9 MHz and 1.9-2.0 MHz, respectively) to increase plasma generation efficiency.

Bias plasma system 580B includes a bias radio frequency ("BRF") generator 531C and a bias matching network 532C. Bias plasma system 580B capacitively couples substrate portion 517 to body member 522, which acts as an auxiliary electrode. The bias plasma system 580B is used to enhance the transport of plasma species (eg, ions) to the substrate surface caused by the source plasma system 580A. In a specific embodiment, as will be discussed below, the bias plasma generator provides up to 10,000 watts of RF power at frequencies less than 5 MHz.

RF generators 531A and 531B include digitally controlled synthesizers and operate over a frequency range between about 1.8 and about 2.1 MHz. As understood by those of ordinary skill in the art, each generator includes an RF control circuit (not shown) that measures the power reflected from the substrate and coil back to the generator and adjusts the operating frequency to obtain the lowest reflected power . RF generators are typically designed to operate with a load having a characteristic impedance of 50 ohms. RF power can be reflected off a load that has a different characteristic impedance than the generator. This can reduce the power delivered to the load. Additionally, power reflected back to the generator by the load can overload and damage the generator. Because the impedance of the plasma ranges from less than 5 ohms to more than 900 ohms depending on the concentration of the plasma, among other factors, and because reflected power can be a function of power, adjusting the frequency of the generator based on reflected power increases from The RF generator delivers the function to the plasma and protects the generator. Another way to reduce reflected power and increase efficiency is to use a matching network.

Matching networks 532A and 532B match the output impedance of generators 531A and 531B to their respective coils 529 and 530 . When the load changes, the RF control circuit can tune the two matching networks by changing the value of the capacitor in the matching network, so that the generator can match the load. The RF control circuit can tune the matching network when the power reflected from the load back to the generator exceeds a certain limit. One way to provide a constant match and effectively prohibit the RF control circuit from tuning the matching network is to set a reflected power limit above any desired value of reflected power. This can help stabilize the plasma under some conditions by keeping the matching network constant at the plasma's nearest conditions.

Other measurements can also help stabilize the plasma. For example, RF control circuitry can be used to determine the power delivered to the load (plasma), and the generator output power can be increased or decreased to keep the delivered power substantially constant during deposition of the layer.

A gas delivery system 533 provides gases to the chamber from several sources 534A-534E through gas delivery lines 538 (only some of which are shown) for processing the substrate. As can be understood by one of ordinary skill in the art, the actual sources for sources 534A-534E and the actual connection of delivery line 538 to chamber 513 vary depending on the deposition and cleaning processes performed within the chamber. Gas is introduced into chamber 513 through gas ring 537 and/or top nozzle 545 . FIG. 5B is a simplified, partial cross-sectional view of chamber 513 showing additional detail of gas ring 537 .

In one embodiment, the first and second gas sources 534A and 534B and the first and second gas flow controllers 535A' and 535B' are fed into the gas ring 537 via gas delivery lines 538 (only some of which are shown). The ring plenum (ring plenum) 536 provides gas. The gas ring 537 has a plurality of source gas nozzles 539 (only one of which is shown for illustration) that provide a uniform gas flow over the substrate. Nozzle length and nozzle angle can be varied to allow uniform distribution and gas utilization efficiency to be tuned for a specific process within a single chamber. In a preferred embodiment, the gas ring 537 has 12 source gas nozzles made of aluminum oxide ceramic.

The gas ring 537 also has a plurality of oxidant gas nozzles 540 (only one of which is shown), which in a preferred embodiment are coplanar with and shorter than the source gas nozzles 539 and at In one embodiment, a plurality of oxidant gas nozzles receive gas from a body plenum 541 . In some embodiments, it is desirable that the source and oxidant gases do not mix prior to injection into chamber 513 . In other embodiments, the oxidant gas and source gas may be mixed prior to injection into the chamber 513 by providing holes (not shown) between the body plenum 541 and the gas ring plenum 536 . In one embodiment, third, fourth and fifth gas sources 534C, 534D and 534D' and third and fourth gas flow controllers 535C and 535D' provide gas to the body plenum through gas delivery line 538. In addition to valves, such as 543B (other valves not shown) can shut off gas from the flow controller to the chamber. In some embodiments of practicing the invention, source 534A includes a source of silane SiH4, source 534B includes a source of molecular oxygen O2, source 534C includes a source of silane SiH4, source 534D includes a source of helium He, and source 534D' includes a source of molecular hydrogen H2.

In embodiments where flammable, toxic or corrosive gases are used, it is desirable to remove the gas left in the gas delivery lines after deposition. For example, this can be accomplished using a 3-way valve such as 543B to isolate chamber 513 from transfer line 538A and communicate transfer line 538A to vacuum foreline 544 . As shown in Figure 5A, other similar valves, such as 543A and 543C, may be incorporated in other gas delivery lines. Such a three-way valve can be placed as close as possible to the chamber 513 to minimize the volume of non-ventilated gas delivery lines (between the three-way valve and the chamber). Additionally, a two-way (on-off) valve (not shown) can be placed between a mass flow controller ("MFC") and the chamber or between the gas source and the MFC.

Referring again to FIG. 5A , the chamber 513 also has a top nozzle 545 and a top vent 546 . Top nozzle 545 and top vent 546 allow independent control of top and side flow of gases, which improves film uniformity and allows fine tuning of film deposition and doping parameters. Top vent 546 is an annular opening around top nozzle 545 . In one embodiment, the first gas source 534A provides a source gas nozzle 539 and a top nozzle 545 . Source nozzle MFC535A' controls the amount of gas delivered to source gas nozzle 539, while top nozzle MFC535A controls the amount of gas delivered to top gas nozzle 545. Similarly, both MFCs 535B and 535B' may be used to control the flow of oxygen from a single oxygen source, such as source 534B, to both top vent 546 and oxidant gas nozzle 540. In some embodiments, oxygen is not provided to the chamber from any side nozzles. The gases provided to top nozzle 545 and top vent 546 may remain separate before flowing into chamber 513 , or the gases may mix within top gas chamber 548 before flowing into chamber 513 . Separate sources of the same gas may be used to supply various parts of the chamber.

A remote microwave generated plasma cleaning system 550 is provided to periodically deposit residues from chamber components. The cleaning system includes a remote microwave generator 551 that generates a plasma from a cleaning gas source 534E (eg, monomolecular fluorine, nitrous oxide, other fluorocarbons, or equivalent) in the reactor cavity 533 . Reactive species generated by this plasma are transported to chamber 513 through applicator tube 555 through cleaning gas input port 554 . Materials used to contain the cleaning plasma (eg, cavity 553 and applicator conduit 555) must be resistant to attack by the plasma. Since the required concentration of plasma species decreases with distance from the reactor chamber 553, the distance between the reactor chamber 553 and the input port 554 should be kept as short as possible. Generating the clean plasma in a remote chamber allows the use of highly efficient microwave generators without subjecting chamber components to the temperature, radiation or bombardment of the glow discharge that occurs in situ in forming the plasma. Accordingly, relatively sensitive components, such as the electrostatic chuck 520, do not need to be covered with dummy wafers or other protection as required in an in-situ plasma cleaning process. In FIG. 5A, plasma cleaning system 550 is shown positioned above chamber 513, although other locations may be used.

A deflector 561 may be provided near the top nozzle to direct the flow of source gas provided in the top nozzle into the chamber, and to direct the flow of remotely generated plasma. Source gas provided by top nozzle 545 is directed into the chamber through a central channel, while remotely generated plasma species provided by cleaning gas input port 554 are directed through baffle 561 to the sides of chamber 513 .

Those skilled in the art will recognize that specific parameters can vary for different processing chambers and different process conditions without departing from the spirit of the invention. Other modifications will also be apparent to those skilled in the art. Such equivalents and modifications are intended to be included within the scope of this invention. The scope of the present invention is therefore not limited to the above-described embodiments, but is defined by the appended claims.

Claims (15)

1. A method of depositing a film on a plurality of substrates, the method comprising: Aging treatment of the processing chamber; sequentially transporting each of the plurality of substrates into the processing chamber to perform a deposition process comprising at least one etch substep on each of the plurality of substrates; and Cleaning the processing chamber between sequentially transferring each of the plurality of substrates, wherein cleaning the processing chamber between sequentially transferring each of the plurality of substrates includes: performing localized cleaning of the processing chamber by forming a high density plasma from a halogen precursor flow; Then, stopping the flow of the halogen precursor; then, heating the processing chamber with plasma from the heating gas; and Cleaning of the processing chamber is then accomplished by forming a high density plasma from the halogen precursor flow. 2. The method according to claim 1, wherein the halogen precursor comprises F2. 3. The method of claim 1, wherein performing a partial cleaning of the processing chamber comprises performing an endpoint partial cleaning of greater than 75% of the cleaning. 4. The method of claim 1, wherein heating the processing chamber comprises: A high-density plasma is formed from the heated gas. 5. The method of claim 4, wherein the heating gas comprises a gas selected from _O2 , Ar and He. 6. The method of claim 4, wherein forming a high density plasma from the heated gas comprises applying source power that is approximately equally divided between a top source and a side source. 7. The method according to claim 1, wherein aging the processing chamber comprises: providing a gas flow of an aged precursor to the processing chamber; forming a high-density plasma from the aged precursor by applying a source power of at least 7500 W, wherein greater than 70% of the source power is distributed at the top of the processing chamber; and Utilize the high density plasma deposition at one point with at least 5000

The thickness of the aging layer. 8. The method of claim 7, wherein the at least 5000

The thickness includes at least 7500

thickness of. 9. The method of claim 7, wherein the at least 5000

The thickness includes at least 10,000

thickness of. 10. The method of claim 7, wherein providing the flow of the aged precursor gas to the processing chamber comprises: providing a flow of silicon-containing gas to the processing chamber; and A gas flow containing oxygen is provided to the processing chamber. 11. The method of claim 10, wherein the flow rate of the oxygen-containing gas is less than the flow rate of the silicon-containing gas. 12. The method of claim 10, wherein the flow rate of the oxygen-containing gas is 0.8 times or less than the flow rate of the silicon-containing gas. 13. The method of claim 10, wherein the silicon-containing gas comprises _SiH4 and the oxygen-containing gas comprises _O2 . 14. The method of claim 10, wherein providing a gas flow of an aging precursor to the processing chamber further comprises providing a gas flow that is non-reactive with the silicon-containing gas and the oxygen-containing gas. 15. The method of claim 14, wherein the gas non-reactive with the silicon-containing gas and the oxygen-containing gas has a flow rate of less than 200 sccm.

Images