US20070076339A1 - Air gap for tungsten/aluminum plug applications - Google Patents

Abstract

An air gap structure substantially reduces undesired capacitance between adjacent interconnects, metal lines or other features in an integrated circuit device. The air gap extends above, and may also additionally extend below, the interconnects desired to be isolated thus minimizing fringing fields between the lines. The integrated air gap structure can be utilized in conjunction with a tungsten plug process. Also, multiple levels of the integrated air gap structure can be fabricated to accommodate multiple metal levels while always ensuring that physical dielectric layer support is provided to the device structure underlying the interconnects.

Description

CROSS REFERENCE TO RELATED APPLICATIONS
• The present application claims priority to and is a continuation of Ser. No. 10/295,080, now U.S. Pat. No. 7,138,329; Ser. No. 10/295,719 filed Nov. 15, 2002; and Ser. No. 11/179,840 filed Jul. 11, 2005. The present application is also related to Ser. No. 10/295,062 filed Nov. 15, 2002 (now U.S. Pat. No. 6,917,109).
• The aforementioned are hereby incorporated by reference as if fully set forth herein.
FIELD OF THE INVENTION
• The present invention relates, in general, to the field of integrated circuit (“IC”) device structures and methods of forming the same. More particularly, the present invention relates to an air gap structure and formation method for reducing undesired capacitive coupling in an integrated circuit device that uses tungsten plugs.
BACKGROUND OF THE INVENTION
• As integrated circuit transistor densities increase, and feature sizes shrink, capacitive coupling between adjacent interconnects, metal lines or other elements also increases. The increased capacitive coupling results in increased parasitic capacitance, which undesirably slows circuit speeds and negatively impacts overall device performance.
• Current attempts to improve electrical isolation in high density integrated circuits involve the implementation of low K dielectric materials such as hydrogen silsesquioxane (HSQ), SiLK™ (a trademark of The Dow Chemical Company) resin, Black Diamond™ (a trademark of Applied Materials company) low K film, Coral™ (a trademark of Novellus System Inc.) carbonaceous oxide film and several other exotic materials. While these materials have a relatively low dielectric constant, they are not normally used in semiconductor manufacturing and therefore increase manufacturing complexity and costs. Much work remains to effectively integrate these materials into conventional semiconductor manufacturing processes.
• Some disadvantages of current low K materials include incompatible thermal coefficient of expansion, low mechanical strength and poor thermal diffusivity.
• Another manner of improving electrical isolation between interconnects is to use an integrated air gap structure because of the extremely low dielectric constant of air. Previous attempts at air gap structures were hard to manufacture and also did not completely isolate adjacent metal lines due to fringing fields above and below the air gap itself.
• For example, U.S. Pat. No. 6,177,329 to Pang (and particularly at col. 7,ll.46+) illustrates one conventional approach in which an additional mask is used to pattern the underlying layers to form the air gaps. This is both inefficient and imprecise for extremely small geometries. U.S. Pat. No. 5,847,439 to Reinberg illustrates another approach in which a combination of a low melting point dielectric, photoresist, a heat cycle and surface tension interact to form a void between two adjacent metal lines. This technique is clearly not suitable for precise control of air gap sizes, and is further disadvantageous because it cannot be used to form gaps which extend above a metal line. The latter may be desirable in some applications. Finally, U.S. Pat. No. 5,949,143 to Bang depicts a rather complex process in which a small opening is made in an etch stop layer and then a selective isotropic etch is used to remove dielectric between two metal lines.
• Clearly, while portions of the aforementioned references are useful in forming air gap structures, and could be used in many applications, their overall approach is not optimal from a manufacturing perspective.
• What is desired, therefore, is an easily manufacturable integrated air gap structure that substantially electrically isolates adjacent interconnects, metal lines or other IC elements.
SUMMARY OF THE INVENTION
• In accordance with the structure and method disclosed herein, a first method for forming a device having an air gap structure includes forming a device layer, which can include first level metal, capacitors, transistors, or other integrated circuit devices, as well as previously formed air gap structures fabricated according to the method of the present invention. A dual damascene structure with a plurality dual damascene opening is formed over the device layer, including first and second patterned dielectric layers. A copper or other conductive layer is formed to fill the dual damascene opening. An adjustable-depth trench is formed between the conductive pattern at least down to the surface of the device layer. The dual damascene structure itself is used as a hard mask in the etching of the trench. Finally, a third dielectric layer is formed onto the trench to form at least one air gap, the air gap optionally extending above the top surface of the dual damascene structure. If desired, the depth of the trench can be extended below the surface of the device layer.
• A second method for forming an air gap structure in an integrated circuit according to the present invention includes forming an interconnect structure on the device layer including, for example, an patterned aluminum or aluminum alloy (conductive aluminum with or without minor amounts of another element or elements) conductive layer overlaying a tungsten conductive plug layer.
• An adjustable-depth trench is formed between the patterned interconnect structure at least down to the surface of the device layer. A dielectric layer is formed over the trench to form an air gap therein, the air gap optionally extending above the top surface of the interconnect structure. If desired, the depth of the trench can be etched to extend below the surface of the device layer.
• A third method for forming an air gap structure for an integrated circuit according to the present invention includes forming an interconnect structure on the device layer including an aluminum alloy interconnect layer overlaying an aluminum alloy plug layer. The conductive plug layer and interconnect layer can be formed simultaneously, thus eliminating at least two processing steps as compared to the second method of the present invention. An adjustable-depth trench is formed between the patterned interconnect structure at least down to the surface of the device layer. A dielectric layer is formed on the trench to form an air gap therein, the air gap optionally extending above the top surface of the interconnect structure. If desired, the depth of the trench can be etched to extend below the surface of the device layer.
• It is an advantage of the present invention that the low dielectric constant of air is used to provide maximum electrical isolation by extending the air gap both below and above the adjacent isolated interconnects, or metal lines, while still ensuring that physical dielectric support is provided beneath the interconnects themselves.
• It is a further advantage of the present invention that the air gap isolation structure is readily manufacturable and compatible with existing semiconductor manufacturing techniques.
• It is a still further advantage of the present invention that exotic low K dielectric materials need not be used, thus saving costs and minimizing manufacturing complexity.
• It is a still further advantage of the present invention that the existence of the air gaps is to release most of the system stress generated by subsequent thermal treatments.
• It is a still further advantage of the present invention that the network structure using conventional dielectric layers encompassing the interconnects provides good thermal dissipation.
BRIEF DESCRIPTION OF THE DRAWINGS
• The aforementioned and other features and objects of the present invention and the manner of attaining them will become more apparent and the invention itself will be best understood by reference to the following description of a preferred embodiment taken in conjunction with the accompanying drawings, wherein:
• FIGS. 1-12 are cross-sectional views of sequential integrated circuit processing steps for forming an air gap isolation structure according to a first embodiment of the present invention, using one of several acceptable dual-damascene metal interconnect processes;
• FIG. 13 is a cross-sectional view of a resulting air gap isolation structure according to the present invention, accommodating the use of multiple levels of a dual-damascene metal interconnect process;
• FIGS. 14-24 are cross-sectional views of sequential integrated circuit processing steps for forming an air gap isolation structure according to a second embodiment of the present invention using one of several acceptable conventional metal interconnect processes;
• FIGS. 25-33 are cross-sectional views of sequential integrated circuit processing steps for forming an air gap isolation structure according to a third embodiment of the present invention using a conventional metal interconnect process;
• FIG. 34 is a cross-sectional view of an air gap isolation structure according to the second/third embodiments of the present invention, accommodating the use of multiple levels of a conventional metal interconnect process; and
• FIGS. 35-38 are cross-sectional views of sequential integrated circuit processing steps for forming an air gap isolation structure according to a fourth embodiment of the present invention, which is a variant of the first embodiment in which an etch stop layer between a line dielectric and a via dielectric is eliminated to further reduce the effective dielectric constant of the inter-metal dielectric layer; and
• FIGS. 39-40 are cross-sectional views of sequential integrated circuit processing steps for forming an air gap isolation structure according to a fifth embodiment of the present invention, which is a variant of the first embodiment in which a first etch is performed only as far as a first etch stop layer.
DESCRIPTION OF PREFERRED EMBODIMENTS
• First Embodiment
• Referring generally now toFIGS. 1-13, a method for forming an integrated circuit device having an air gap structure is shown for a dual damascene-type metal interconnect structure.
• InFIG. 1, adevice layer10 is formed, which may be a simple silicon substrate and first-level metal, for example. Thedevice layer10 may nonetheless also include multiple levels of metal, transistors, capacitors, or other devices, including previously manufactured integrated air gap structures built according to the method of the present invention. Thus,device layer10 is meant to represent that portion of the previously formed integrated circuit device on which the air gap structure is to be built, but it is not limited to any particular form, structure or circuitry.
• Similarly, as used herein, the terms “on” or “onto” or “above” when used in connection with various thin film layers are merely intended to denote a physical spatial relationship, and not necessarily a direct physical or electrical contact. It will be understood therefore by those skilled in the art that in embodiments of the invention, a first layer may be “on” or “above” a second layer, even if there are other intervening layers present.
• In a first embodiment, a firstetch stop layer12 is formed on the upper surface of thedevice layer10. Theetch stop layer12 is ideally formed of silicon nitride (SiNx), silicon oxynitride (SiNxOy), silicon carbide (SiCx), or the like, and is deposited to a thickness of about 100 to 1500 Angstroms using any of a number of known conventional mechanisms. The particular material for any application of course can be determined by one skilled in the art by coordinating such selection with an etch chemistry/mechanism to be employed in a later etch operation. Thus, so long as such first etch stop layer is otherwise compatible with other materials and processes described herein, the present invention is not limited to any particular material.
• A first dielectric layer14 (designated generally herein as a “via” dielectric layer because the body of a via contact is later formed therein) is formed onetch stop layer12. Thefirst dielectric layer14 is ideally silicon dioxide or undoped silicate glass (USG) but can also be fluorinated silicate glass (FSG), or borophosphorus silicate glass (BPSG), phosphorus silicate glass (PSG), or the like and is deposited to a thickness of about 1000 to 10000 Angstroms using well-known processing tools. Moreover,first dielectric layer14 can include combinations and/or composites of individual thin film layers. Again, the particular formulation for this layer will depend on desired performance characteristics and process requirements, and thus a variety of materials are expected to be suitable for such layer.
• InFIG. 2, an additional secondetch stop layer16 is deposited onto the upper surface of viadielectric layer14. As withetch stop layer12, the particular composition ofetch stop layer16 is not critical, and can be determined without undue experimentation by one skilled in the art based on the present teachings and objectives defined herein for the inventions presented.
• A second dielectric layer18 (designated generally herein as a “line” dielectric layer because portions of a conductive line are later formed therein) is deposited onto the surface ofetch stop layer16. The line dielectric is also ideally silicon dioxide or a similar dielectric as viadielectric layer14 and is deposited to a thickness of about 1000 to 10000 Angstroms. The selection of materials for this layer will again be a routine design choice based on lithographic and etching requirements associated with a particular manufacturing process.
• A third etch stop and/or ananti-reflecting layer20 is subsequently deposited on theline dielectric layer18. Etch stop and/oranti-reflecting layer20 is preferably SiNx, SiNxOy, silicon riched oxide (SRO), SiCx or the like and is deposited to a thickness of about 100 to 500 Angstroms. As with the other etch stop layers, the particular material for any application of course can be determined by one skilled in the art by coordinating such selection with an etch chemistry/mechanism to be employed in a later etch operation.
• In general, the overall composition of the structure shown inFIG. 2 can be constructed with conventional and well-known manufacturing equipment suitable for wafer processing operations. The particular selection of materials for the thin film layers is directed primarily by concerns of reliability, reproducibility and lithographic constraints in small scale geometries, and so it is expected that a wide variety of combinations will be suitable for use in the present invention.
• InFIG. 3, aphotoresist layer22 is formed on third etch stop and/oranti-reflecting layer20 to a thickness of about 1000 to 10000 Angstroms.Photoresist layer22 is patterned to form metal contacts or a viapattern24A by any conventional photolithography process. The particular resist formulation and lithography process are again not material to the present teachings, so any suitable combination may be employed.
• InFIG. 4,line dielectric layer18 is anisotropically etched using viapattern24A as a mask to form a metal contact or viaopening24B. A conventional oxide etch such as reactive ion etch (RIE) can be used for this step, which is terminated upon reaching firstetch stop layer12, or some other point before this. Other techniques will be apparent to those skilled in the art. It should be noted, of course, thatetch stop layer12 can also be removed in those areas (not shown) where it may be desirable to make a conductive contact to some portion ofdevice layer portion10.
• InFIG. 5, after removing the resistlayer22, anotherphotoresist layer32 is processed to form ametal line pattern24C by photolithography processes. Again, the particular resist formulation and lithography process forlayer32 are again not material to the present teachings, so any suitable combination may be employed.
• InFIG. 6, bothline dielectric layer18 and viadielectric layer14 are etched to form anopening24D for subsequent processing of the dual damascene structure. This etching operation is also done with a conventional etch such as reactive ion etch (RIE) can be used for this step, which is preferably terminated upon reaching secondetch stop layer16. Other techniques will be apparent to those skilled in the art. Thus, bothphotoresist layer32 and the patternedetch stop layer16 act as a form of mask for this operation.
• It should be noted that the upper portion of opening24D serves as an interconnect line while the bottom portion ofopening24D functions as a conductive pillar to the device portion. The result is aconductive line28 with a cross section in some areas that resembles a T-shape as seen in the Figures.
• InFIG. 7, resistlayer32 is stripped using a conventional process and a composite copper barrier/seed layer (shown as a singleintegrated layer26 for simplicity) is deposited using conventional means. The first portion of copper barrier/seed layer is a barrier layer selected from a group of conductive materials that can prevent Cu from diffusing into adjacent dielectric layers, such as Ta, TaN, TiN, TiW, WN, Mo, W, etc. These are examples known to the inventors at this time, and it is possible of course that later developed materials unforeseen and as yet undiscovered may prove to be suitable for this purpose.
• A seed layer portion of composite barrier/seed layer26 is typically Cu or Cu alloy, again deposited using known means.
• In a preferred embodiment, the copper barrier layer portion is deposited to a thickness of about 50 to 500 Angstroms, and the seed layer portion is deposited to a thickness of about 300 to 2000 Angstroms to form combinedlayer26. It will be understood by those skilled in the art that these values are merely exemplary for the geometries described therein, and that the final values for any particular embodiment of the invention may deviate from such figures.
• InFIG. 8, opening24D is then filled with acopper layer28. Copper is deposited to a thickness of about 2000 to 10000 Angstroms using any well-known conventional tools, which preferably completely fills opening24D and provides an excess copper layer. It will be understood, of course, that the deposition of this layer may be achieved in a single step, or multiple steps to provide a graded and/or composite copper layer withinopening24D.
• InFIG. 9, any excess copper on top ofline dielectric18 is removed preferably using chemical-mechanical polishing (CMP) with a suitable polish pad, slurry, recipe, etc. as is known to those skilled in the art. In self-limiting growth processes, this type of CMP operation may be minimized or reduced. The above steps for defining the openings and forming the Cu lines withinsuch openings24D are merely an example of the preferred technique known to the inventors at this time, and it is possible of course that later developed processes unforeseen and as yet undiscovered may prove to be suitable for such purposes.
• InFIG. 10, a plurality of dual damascene metalconductive lines28 form aninterconnect structure28′. Each dual damascenemetal interconnect line28 is isolated primarily at this point by a combination ofdielectric layers14 and18.
• Other cross-sectional portions of a wafer are illustrated inFIG. 10A to show some additional examples of structures/relationships that may exist. For example, in some areas anconductive line29 may not extend down to device layer10 (the most likely case for a metal line); inother areas29′ the position of the via is not symmetric about the metal line. Inother areas29″ the via part of the dual damascene structure may extend to thedevice portion10 and may be contacted to the substrate. Inother area29′″, the metal line part of the dual damascene structure is about the same width as that of via parts. Thus, a variety of cross-sectional patterns will result. It will be understood by those skilled in the art that these are merely exemplary, and that other portions of a wafer are likely to contain additional variants of those illustrated depending on interconnect/masking requirements.
• As alluded to earlier, at least some of theconductive lines28 may be included as part of a so-called “dummy” pattern so as to make the interconnection patterns more uniform across the surface of a wafer. This also facilitates the manufacturing process because the resulting surface is more uniform.
• InFIG. 10B, a side perspective can be seen of another exemplaryconductive line28 viewed lengthwise as it may be formed for an integrated circuit. At individual points across the surface, a lower portion ofconductive line28 extends (in some instances) as a type ofconductive pillar11 to form an electrical contact at selective points todevice layer10. These conductive pillars are formed from a combination of material fromconductive line28 that is surrounded bydielectric material14 for additional support.
• InFIG. 11,dielectric layers14 and18 fromFIG. 10 are preferably anisotropically etched usingcopper layer28 as a hard mask. A conventional dielectric etch is used to formtrenches30 intodielectric layers14 and18. The form and depth oftrenches30 is adjustable and can extend down to the upper surface of thedevice layer10, or can be etched further to extend down belowetch stop layer12 and below the surface of device layer10 (not shown inFIG. 11).
• For reasons that are explained in more detail below, an anisotropic etch (or an etch type with reduced isotropic behavior) is preferred over a “wet” isotropic etch at this point, because it is desirable to leave some small amount of dielectric on the sidewalls ofinterconnect18, underneath the overhang areas as seen inFIG. 11. Of course, in some cases it may be desirable to remove such remaining material (from layer14) and replace it with another material (i.e., through another spin on deposition/plasma deposition and subsequent etch. An isotropic etch could then be used onlayer14. While this would require additional processing steps, it is conceivable that the dielectric constant could be improved in this fashion, as well as reliability, yield etc. of the overall process.
• The depth oftrenches30 is preferably controlled through a timed etch, and it will be apparent to those skilled in the art that the duration of such etch will be a function of the dielectric layer composition, the etch process chemistry, the thickness oflayers14,18, etc., etc. The etch time will thus vary from application to application, and can be determined with routine simulations and testings.
• Alternatively it is possible instead to use eitheretch stop layer12 to control the end of the etch, and/or to provide yet another etch stop layer (not shown) withinlayer14 at any optimally determined etch depth. In such instance, of course,layer14 would be a composite layer deposited in separate steps, and thus this option is not as attractive from a throughput perspective.
• As noted above, a preferred approach uses copperconductive lines28 as a mask, but it those skilled in the art will appreciate that an additional masking step could be employed should it be necessary to make the air gaps more narrow. Again, this is not optimal from a control and throughput perspective, so it is probably not desirable except in limited cases.
• In contrast, in the present invention, it should be relatively simple and easy to control the size of such air gaps both by controlling the spacing between theconductive lines28, as well as tailoring the size/shape of the top portion of the conductive line. This is true since the latter effectuate the hard mask used for etchingdielectric layers14,18 to form the air gaps.
• In this respect, those skilled in the art will appreciate that shapes and sizes of the interconnect structures shown in the figures are only approximate, and not intended to be to scale. Other variations are expected to be beneficially employed in accordance with the present teachings.
• InFIG. 12, acopper barrier layer44 such as SiNx, SiC, or the like is deposited to a thickness of about 50 to 500 Angstroms. Again, these are materials particularly suited for copper, and other compositions may be needed for other types of conductive line metals. For some metals, of course, a barrier layer may not be needed in the first place.
• A silicon-dioxide dielectric layer, or the like32 is then deposited to a thickness of about 2000 to 10000 Angstroms. Poor step coverage by the deposition ofdielectric layer32, such as conventional plasma enhanced chemical vapor deposition (PECVD), results in the formation of intra-metalline air gaps34. In other words, the present invention exploits the basically conformal growth nature of this type of process to intentionally form gaps between the metal lines. By controlling the deposition parameters, and the thickness of the deposited layer, the size, shape and height ofair gaps34 can be customized for any particular line interconnect geometry.
• In lieu of a PECVD process, other similar techniques that are characterized by poor step coverage could be used to formair gaps34. For example, a series of HDPCVD depositions could be used. As those skilled in the art will appreciate, the above are merely examples of techniques for achieving poor step coverage that are known to the inventors at this time, and it is possible of course that later developed processes unforeseen and as yet undiscovered may prove to be suitable for such purposes.
• As previously discussed, the inclusion ofair gaps34 provides superior electric isolation due to the low dielectric constant of air. The size and shape ofair gaps34 may also vary across the surface of a wafer, as illustrated generally inFIGS. 12A. It can be seen in such picture that the width of any air gaps (W1 or W2) are not necessarily uniform across the surface of the wafer, nor are they required to be for purposes of the present invention. It is simply desirable, of course, to ensure that at least some air filled gap is provided between two adjacent signal lines.
• Thus, as seen inFIG. 12A, one useful benchmark is to consider the relative ratio of the airgap width (W1, W2) to an overall line spacing (WSI, WS2). In general, the closer W1/WS1 and W2/WS2 are to unity, the lower the capacitance, so it is preferable to maximize this value to the extent consistent with other processing requirements.
• In addition, the height by which theair gaps34 extend aboveinterconnect layer28, or below such layer, is controlled both by the trench sizing noted earlier, as well as the details of the conformal dielectric deposition noted earlier. Thus, they may also vary in vertical size as seen inFIG. 12B, where two different heights (H1 and H2) are provided. Again it is understood that the height of any air gaps (H1 or H2) are not necessarily uniform across the surface of the wafer, nor are they required to be for purposes of the present invention. Nonetheless, for reasons well understood in the art, it is preferable (to the exent possible within available process constraints) to maximize such air gap heights (in relation to the height HL of the conductive lines28) by extending them above and below aninterconnect structure28 to reduce the capacitance between adjacent lines.
• In summary, an inter-line interconnect structure as shown inFIG. 12 typically includes ametal line28, an conductive linesidewall dielectric portion14′, a seconddielectric filler32, andair gap34. The sidewalldielectric portion14′ left underneathmetal line28 provides structural support and additional process window margin when the present invention is used in small scale line width geometries.
• Those skilled in the art will further appreciate that the above are merely examples of what might be present in any section of the wafer, and that other air gap structures will inevitably result as part of any conventional manufacturing process employing the present teachings.
• As further noted, to reduce non-uniformities for such air gaps, dummy metal lines can be added to an interconnect pattern to ensure that no large flat spaces are left between adjacent conductive lines. Thus, for example, inFIG. 12, for some instances across the surface of the wafer, themiddle metal line28 may be carrying an actual signal, and in other instances, a “dummy”metal line28 may be simply added so as to create a uniform capacitance everywhere for the metal lines adjacent thereto.
• InFIG. 13, a composite drawing is shown of two dual damascene structures fabricated in sequence according to the method of the present invention. Adevice layer10 includes a silicon substrate and a portion of first level ofmetal interconnect28 extends herein as well. In a first level of interconnect structure according to the present invention,copper metal layer28 andair gaps34 are shown extending in and to the top ofdielectric layers14 and18. Note thatair gaps34 are shown to extend below the level of the upper surface of thedevice layer10 as well as above the upper surface ofmetal lines28, thus providing the maximum electrical isolation between adjacent metal structures.
• Also shown inFIG. 13 is a second level of metal interconnect that includes anadditional metal interconnect38 and adielectric layer36 withair gaps40 and42.Air gap40 provides intra-level metal isolation and extends to the surface of devicelayer including layers10,14, and18, as well as above the upper surface ofmetal lines38.Air gaps42 extend below the surface of the devicelayer including layers10,14, and18, and thus provide electrical intra-metal isolation for bothmetal layers38 and28.
• Furthermore it will be apparent that this overall process could be repeated as needed to form additional interconnect layers, and the present invention is by no means limited to any particular number of such layers.
• Another important observation about the present invention that can be gleaned fromFIG. 13, is that in some instances an air gap for a second level interconnect may be formed on top of a first level interconnect. In other instances a single air gap can be extended in height so that it serves to reduce capacitance for more than one interconnect layer. For example, theair gap42 shown in the middle ofFIG. 13 serves as an air gap for two separate metal interconnect levels; this same principle could be extended as needed for additional levels. Thus by appropriate “stacking” and arrangement of interconnect layers, a single air gap can be formed between adjacently located conductive lines in more than one layer of metal.
• As illustrated herein, thedielectric material14 underneath the conductive lines further functions to provide some measure of structural support for the latter. This feature can be enhanced or reduced in other embodiments by structural variations so that more or less dielectric is left on the sidewalls, or under the top portions of the conductive lines. The dielectric also functions as a heat dissipator, and further reduces electromigration. Accordingly, the amount of dielectric left on the sidewalls can be tailored for any particular environment, so that it might be used extensively in some applications (thicker layers), and not used in others (thin layers, or no layers at all).
• Second Embodiment
• Referring generally now toFIGS. 14-24, a method for forming an integrated circuit device having at least one air gap structure is shown for a conventional metal interconnect structure of the type having aluminum alloy metal interconnect layers and tungsten metal plugs. Except where otherwise noted, like numerals are intended to represent like structures and materials already identified in connection withFIGS. 1-13.
• InFIG. 14, adevice layer10 is formed as before.
• A contact/viadielectric layer14 is formed ondevice layer10. As before,dielectric layer14 is ideally silicon dioxide but can also be USG, FSG, PSG, BPSG, or the like and is deposited to a thickness of about 1000 to 10000 Angstroms. It will be understood, of course, thatlayer14 may be comprised of a combination of layers, and formed in more than one processing step, but for purposes of the present discussion, it will be referred to as a single layer.
• InFIG. 15, aphotoresist layer22 is formed ondielectric layer14 to a thickness of about 1000 to 10000 Angstroms.Photoresist layer22 is patterned to form metal contact or viapattern56A by photolithography processes as before.
• InFIG. 16, as noted before,openings56B are etched into the contact/viadielectric layer52 in a similar fashion to that already described forFIG. 4.
• InFIG. 17, resistlayer22 is stripped and a tungsten barrier layer92 (such as Ti/TiN, Ta, TaN etc.) is deposited on the surface ofdielectric layer52 and in openings56. Again, these are merely examples of those known at this time to be particularly suited for Tungsten, and other compositions may be needed for other types of conductive line metals. For some metals, of course, a barrier layer may not be needed in the first place.
• A layer ofTungsten58 is then preferably deposited to a thickness of about 500 to 8000 Angstroms, which completely fills openings56. Again, for other processes, materials other than Tungsten may be more suitable.
• InFIG. 18, any excess tungsten is removed using tungsten CMP or tungsten etch back, which results in a structure that includesdielectric layer14 and tungsten metal plugs58. In self-limiting growth processes, this type of CMP operation may be minimized or reduced.
• InFIG. 19, an aluminum alloy (or the like)interconnect layer60 is deposited on combined metal plug/dielectric layer14/58 to a thickness of about 2000 to 10000 Angstroms. Again, for other processes, materials other than an aluminum alloy may be more suitable. For example, doped polycrystalline silicon is also well-known as an effective conductive interconnect/gate material.
• InFIG. 20, a resistlayer62 is formed on themetal layer60 in any conventional manner to a thickness of preferably about 2000 to 15000 Angstroms and followed preferably by a photolithography process to result inmetal line pattern64A.
• InFIG. 21, anintra-metal spacing64B is formed by etchingmetal layer60 using a conventional metal etching process to form an interconnect structure consisting of patternedmetal layer60 andspacings64B. Again, the particular etch chemistry and technique will depend on the particular material selected forlayer60.
• InFIG. 22, the metal interconnect structure ofFIG. 21 is shown in conjunction with several other metal plugs58, each capped by a section ofmetal interconnect layer60. It will be understood, of course, that it is not necessary to locate every interconnect line above a metal plug.
• In a preferred first processing option, any material inspacings64B is removed and etched down to the surface of thedevice layer50 with the resistlayer62 intact to formtrenches64C. As explained in connection withFIG. 11 as well, the depth oftrenches64C is adjustable and can be made down to and even below the upper surface of the device layer10 (not shown inFIG. 22).
• In a second processing variation of this embodiment (shown inFIG. 23), resistlayer62 is first stripped and previously etchedmetal layer60 is used as a hard mask to etchtrenches64C. The choice between these two variations can be made on a case by case basis in accordance with conventional and well-known process requirements.
• InFIG. 24 air gaps are formed in substantially the same manner as depicted earlier forFIG. 12. That is, a silicon-dioxide or thelike dielectric layer66 is deposited onto the surface to a thickness of about 2000 to 10000 Angstroms. Poor step coverage by the deposition ofdielectric layer66 results in the formation of intra-metalline air gaps68.Air gaps68 provide superior electric isolation due to the low dielectric constant of air as previously discussed.
• It will be appreciated by those skilled in the art that this second embodiment can also be used to create structures that are similar to those already illustrated inFIGS. 12A and 12B, including air gaps of different height, width, etc. Moreover, the above steps can be sequenced again to form multi-level interconnect structures in the same manner as previously described forFIG. 13. Thus, air gaps can be used as an insulation layer between inter-metal or intra-metal layers formed of Al, Al alloys, polycrystalline silicon, etc.
• Third Embodiment
• Referring generally now toFIGS. 25-33, a third embodiment of a method for forming an integrated circuit device having at least an air gap structure is shown for a conventional metal interconnect structure of the type having aluminum alloy metal interconnect layers and aluminum alloy metal plugs. The primary difference to the second embodiment is in the use of a different type of a barrier metal layer for the interlayer plugs.
• InFIG. 25, a contact/viadielectric layer14 is formed ondevice layer10 as before.
• InFIG. 26, aphotoresist layer22 is formed and patterned ondielectric layer14 as before to form a pattern ofopenings86A.
• InFIG. 27,openings86B are etched into contact/viadielectric layer14 as before.
• InFIG. 28, resistlayer22 is stripped and an aluminum barrier layer94 (such as Ti/TiN, Ta, TaN or Aluminum oxide) is deposited on the surface ofdielectric layer82 and inopenings86. Again, these are merely examples of those known at this time to be particularly suited for Aluminum, and other compositions may be needed for other types of conductive line metals. For some metals, of course, a barrier layer may not be needed in the first place.
• An aluminum alloy layer90 (preferably Aluminum with some small percentage of Cu and/or Si) is then deposited to a thickness of about 500 to 8000 Angstroms, which completely fills contact/viaopenings86B and provides an aluminum alloy interconnect layer coupled to aluminum alloy plugs88.
• This embodiment, therefore, is distinguished from the second embodiment noted earlier in that the plug and interconnect layer can be formed in a single step, thus improving throughput for those applications where it is acceptable to use something other than a Tungsten based plug.
• InFIG. 29, as before a resistlayer92 is formed on the metal layer to a thickness of about 2000 to 15000 Angstroms followed by a photolithography process.
• InFIG. 30, an intra-metal spacing74B is formed by etching thealuminum metal layer90 using a conventional metal etching process as noted earlier forFIG. 21.
• InFIG. 31, the metal interconnect structure ofFIG. 30 is shown in conjunction with several other metal plugs88, each capped by a section of aluminum alloymetal interconnect layer90. As before, it will be understood, of course, that it is not necessary to locate every interconnect line above a metal plug.
• In a preferred first processing option, any material in intra-metal spacings74B is removed and etched down to the surface of thedevice layer10 with the resistlayer92 intact to form trenches74C. As explained in connection withFIG. 11 as well, the depth of trenches74C is adjustable and can be made down to and even below the upper surface of the device layer10 (not shown inFIG. 31).
• In a second processing variation shown inFIG. 32, resistlayer92 is first stripped and previously etchedmetal layer90 is used as a hard mask to etchtrenches74. Again the choice between these two variations can be made on a case by case basis in accordance with conventional and well-known process requirements.
• InFIG. 33 air gaps are formed in substantially the same manner as depicted earlier forFIG. 12. That is, a silicon-dioxide dielectric layer or the like78 is deposited to fill thetrenches74 and cover themetal pattern90 to a thickness of about 1000 to 8000 Angstroms. Poor step coverage by the deposition ofdielectric layer78 results in the formation of intra-metalline air gaps76.Air gaps76 provide superior electric isolation due to the low dielectric constant of air as previously discussed.
• It will be appreciated by those skilled in the art that this third embodiment can also be used to create structures that are similar to those already illustrated inFIGS. 12A and 12B, including air gaps of different height, width, etc.
• Moreover, the above steps can be sequenced again to form multi-level interconnect structures in the same manner as previously described forFIG. 13, and as shown generally inFIG. 34.
• InFIG. 34, a composite drawing is shown of two metal interconnect structures fabricated according to the third (and second) air gap method of the present invention. In a first level of interconnect structure according to the present invention,metal interconnect60, metal plugs58, andair gaps68 are shown embedded indielectric layer14 abovedevice layer10.
• Note that as withFIG. 13, the resulting structure ofFIG. 34 shows that thatair gaps68 can extend below the level of the upper surface ofdevice layer50 and above the upper surface ofmetal lines60, thus providing maximum electrical isolation. Also shown inFIG. 34 is a second level ofmetal interconnect layer80 that includes anadditional metal level86, metal plugs84, andair gaps88.Air gap88 provides intra-level metal isolation and extends tolayers50,52, and60.
• Fourth Embodiment
• A fourth embodiment is now described with reference toFIGS. 35-38. This embodiment is a variant of the first embodiment in which an etch stop layer between a line dielectric and a via dielectric is eliminated to further reduce the effective dielectric constant of the inter-metal dielectric layer.
• Thus, inFIG. 35, the secondetch stop layer16 between the via and linedielectric layers14 and18 (FIG. 2) has been eliminated to further reduce the effective dielectric constant of the inter-metal dielectric layer. In lieu of two dielectric layers separated by an etch stop layer, asingle dielectric layer15 is deposited onto the surface ofetch stop layer12. Thesingle dielectric layer15 is also ideally silicon dioxide or the like and is deposited to a thickness of about 1000 to 10000 Angstroms in a manner similar to that already described for viadielectric layer14.
• An etch stop and/oranti-reflecting layer20 is subsequently deposited on theline dielectric layer15 as discussed before in connection withFIG. 2.
• InFIG. 36, aphotoresist layer22 is formed on etch stop and/oranti-reflecting layer20 to a thickness of about 1000 to 10000 Angstroms as already described inFIG. 3.
• InFIG. 37, using metal contact or viapattern24A as a mask, thedielectric layer15 is anisotropically etched to form metal contact or viaopening24B as already described inFIG. 4. The primary difference fromFIG. 4 is that, as generally illustrated,dielectric layer15 is only partially etched, in this case, to a depth of approximately slightly more than half the thickness of such layer.
• InFIG. 38, after removing the resistlayer22, anotherphotoresist layer32 is processed to formmetal line pattern24C by photolithography processes as generally already described inFIG. 5.
• From this point forward, processing takes place in substantially the same fashion as already illustrated above in connection withFIGS. 6-13, thus resulting in a single or multi-level air gap interconnect structure, albeit with slightly modified layer compositions as noted here.
• Fifth Embodiment
• A fifth embodiment is now described with reference toFIGS. 39-40. This embodiment is also a variant of the first embodiment in which a first etching operation is performed only as far as a first etch stop layer.
• Accordingly,FIG. 39 illustrates a variation in which given the structure shown inFIG. 3, an etching operation is conducted in a similar fashion to that already describe inFIG. 4, except that such etch is stopped upon reaching secondetch stop layer16. In all other respects, this operation is the same, in thatline dielectric layer18 is anisotropically etched using viapattern24A as a mask to form a metal contact or viaopening24B. It is only the case, therefore, that these openings do not extend as far down as those illustrated inFIG. 4.
• InFIG. 40, after removing the resistlayer22, anotherphotoresist layer32 is formed. A subsequent etch transfers the upper profile of opening24C to the bottom of the openings, so that a deeperenlarged opening24D results that is substantially the same as shown inFIG. 6.
• From this point forward, processing takes place in substantially the same fashion as already illustrated above in connection withFIGS. 7-13, thus resulting in a single or multi-level air gap interconnect structure, albeit with slightly modified layer compositions as noted here.
• While there have been described above the principles of the present invention in conjunction with specific circuit implementations and applications it is to be clearly understood that the foregoing description is made only by way of example and not as a limitation to the scope of the invention. Particularly, it is recognized that the teachings of the foregoing disclosure will suggest other modifications to those persons skilled in the relevant art. Such modifications may involve other features which are already known and which may be used instead of or in addition to features already described herein. Although claims have been formulated in this application to particular combinations of features, it should be understood that the scope of the disclosure herein also includes any novel feature or any novel combination of features disclosed either explicitly or implicitly or any generalization or modification thereof which would be apparent to persons skilled in the relevant art, whether or not such relates to the same invention as presently claimed in any claim and whether or not it mitigates any or all of the same technical problems as confronted by the present invention. The applicants hereby reserve the right to formulate new claims to such features and/or combinations of such features during the prosecution of the present application or of any further application derived therefrom.

Claims (12)

1-36. (canceled)

37. An air gap structure in an integrated circuit comprising:

a conductive structure within a first dielectric layer on a substrate, said conductive structure including conductive plugs;

an opening on the first dielectric layer and adjacent to the conductive structure and

a second dielectric layer situated over the substrate to form at least an air gap in the opening.

38. An air gap structure in an integrated circuit comprising:

a device layer;

a first conductive line and a second conductive line situated above said device layer;

said first conductive line including a first conductive plug, and said second conductive line including a second conductive plug;

a trench located between said first conductive line and a second conductive line adjacent to said first conductive plug and said second conductive plug;

a first dielectric layer arranged so as form sidewalls of said trench between said first conductive line and a second conductive line and adjacent to said first conductive plug and second conductive plug;

a second dielectric layer situated over said first conductive line, said second conductive line and said trench;

wherein an air gap is situated between said first conductive line and a second conductive line and is defined by a region bounded by a bottom of said trench, said sidewalls, said first conductive line and said second conductive line, and said second dielectric layer.

39. The air gap structure ofclaim 38 in which said air gap comprises an air gap extending above a top surface of said first conductive line.

40. The air gap structure ofclaim 38 wherein both of said plugs extend and electrically couple to said device layer.

41. The air gap structure ofclaim 38 wherein said first conductive plug and second conductive plug are selected from tungsten or aluminum based metal.

42. The air gap structure ofclaim 38 wherein said conductive lines are made of aluminum or aluminum alloy layer.

43. The air gap structure ofclaim 38 in which said trench extends below a top surface of said device layer.

44. The air gap structure ofclaim 38 in which said second dielectric layer comprises a dielectric layer selected from the group including SiO2, USG, PSG, FSG and BPSG.

45. The air gap structure ofclaim 44 in which said second dielectric layer comprises a dielectric layer about 1000 to 10000 Angstroms thick.

46. An integrated circuit comprising:

a device layer;

a plurality of conductive lines situated above said device layer;

a plurality of conductive plugs coupling said conductive lines to said device layer;

a plurality of trenches located between adjacent ones of said plurality of conductive lines;

a first dielectric material arranged so as to line at least portions of sidewalls of at least some of said plurality of trenches;

a second dielectric layer situated over said plurality of conductive lines and said plurality of trenches;

wherein a plurality of air gaps are situated between said adjacent ones of said plurality of conductive lines, said plurality of air gaps being defined by separate spaces bounded by bottoms of said plurality of trenches, said sidewalls, and said second dielectric layer;

said plurality of air gaps being further arranged such that at least some of said plurality of air gaps extend below a top surface of said device layer, and at least some of said plurality of air gaps extend above a top surface of said plurality of conductive lines.

47. An integrated circuit comprising:

a device layer;

an interconnect structure including a first plurality of conductive lines situated at a first level above said device layer, and a second plurality of conductive lines situated at a second level above said device layer;

wherein said first plurality of conductive lines are coupled to said device layer through a first level of one or more conductive plugs, and said second plurality of conductive lines are coupled to said first plurality of conductive lines through a second level of one or more conductive plugs;

wherein said first plurality of conductive lines and said second plurality of conductive lines, and/or said first level of one or more conductive plugs and said second level of one or more conductive plugs overlap in regions of the integrated circuit resulting in a third plurality of multi-level conductive lines;

a plurality of trenches located between adjacent ones of said third plurality of multi-level conductive lines;

a first dielectric material arranged so as to line at least portions of sidewalls of at least some of said plurality of trenches;

a second dielectric layer situated over said third plurality of multi-level conductive lines and said plurality of trenches;

wherein a plurality of multi-level air gaps are situated between said adjacent ones of said third plurality of conductive lines, said plurality of multi-level air gaps being defined by separate spaces bounded by bottoms of said plurality of trenches, said sidewalls, and said second dielectric layer;

said plurality of air gaps being further arranged such that at least some of said plurality of air gaps extend below a top surface of said device layer, and at least some of said plurality of air gaps extend above a top surface of said third plurality of multi-level conductive lines.

Images