KR20050111391A - Slurry compositions for use in a chemical-mechanical planarization process having non-spherical abrasive particles - Google Patents

Abstract

A chemical-mechanical abrasive composition for use in semiconductor processing uses abrasive particles having a non-spherical morphology.

Description

Slurry composition with non-spherical abrasive particles for use in a chemical-mechanical planarization method {SURRY COMPOSITIONS FOR USE IN A CHEMICAL-MECHANICAL PLANARIZATION PROCESS HAVING NON-SPHERICAL ABRASIVE PARTICLES}

The present invention relates to novel slurries for chemical-mechanical planarization (CMP). The present invention is applicable to fabricating high speed integrated circuits with submicron design shapes and high conductivity internal interconnect structures with high production throughput.

In the fabrication of integrated circuits and other electronic devices, multiple layers of conductive, semiconducting and dielectric materials are deposited on or removed from the surface of the substrate. Thin layers of conductive, semiconductive and dielectric materials can be deposited by a number of deposition techniques. Conventional deposition techniques in modern processes include physical vapor deposition (PVD), chemical vapor deposition (CVD), plasma-enhanced chemical vapor deposition (PECVD), and also known as sputtering. Electrochemical plating (ECP) is included.

Since the layers of material are deposited and removed sequentially, the top surface of the substrate may be non-planar across its surface and may require planarization. Planarizing or "polishing" the surface is the process of removing material from the surface of the substrate to form a generally flat planar surface. Planarization is useful for removing undesirable surface topography and surface defects such as rough surfaces, aggregated materials, crystal lattice damage, scratches and contaminated layers or materials. Planarization is also useful for shaping the shape on the substrate by removing excess deposits used to fill the shape and provide a flat surface for subsequent metallization and processing levels.

Chemical mechanical planarization or chemical mechanical polishing (CMP) is a common technique used to planarize a substrate. CMP utilizes chemical compositions, generally slurries or other fluid media, for selectively removing material from the substrate. Considerations in CMP slurry design have been reviewed in Rajiv K. Singh et al., "Fundamentals of Slurry Design for CMP of Metal and Dielectrics Materials", MRS Bulletin, pages 752-760 (October 2002). In conventional CMP techniques, a substrate carrier or polishing head is mounted on a carrier assembly and placed in contact with a polishing pad within a CMP apparatus. The carrier assembly provides adjustable pressure on the substrate to propel the substrate against the polishing pad. The pad is moved relative to the substrate by the external driving force. Thus, CMP devices cause polishing or rubbing movement between the surface of the substrate and the polishing pad, while dispersing the polishing composition or slurry resulting in both chemical and mechanical activity.

Conventional slurries used in the CMP process contain abrasive particles in a reactive solution. Alternatively, the abrasive may be a fixed abrasive such as a fixed abrasive polishing pad, which may be used with a CMP composition or slurry that does not contain abrasive particles. Fixed abrasive products generally comprise a backing sheet to which a plurality of geometric abrasive composite elements are attached.

The most widely used abrasives in semiconductor CMP methods are described in U.S. Patent Nos. 4,959,113; 5,354,490; And 5,516,346, and WO 97 / 40,030, which may be produced by a fuming or sol-gel method, such as silica (SiO ₂ ), alumina (Al ₂ O ₃ ), ceria (CeO ₂ ), zirconia (ZrO ₂ ) and titania (TiO ₂ ). Recently, compositions or slurries containing manganese (Mn ₂ O ₃ ) (European Patent No. 816,457) or silicon nitride (SiN) (European Patent No. 786,504) have been reported.

U.S. Pat.No. 6,508,952 discloses abrasives in the form of particles that are commercially available, such as SiO ₂ , Al ₂ O ₃ , ZrO ₂ , CeO ₂ , SiC, Fe ₂ O ₃ , TiO ₂ , Si ₃ N _4, or mixtures thereof. The containing CMP slurry was described. These abrasive particles typically have high purity, high surface area and narrow particle size distribution, and are therefore suitable for use as abrasives in abrasive compositions.

US Pat. No. 4,549,374 describes polishing a semiconductor wafer using an abrasive slurry prepared by dispersing montmorillonite clay in deionized water. The pH of the slurry is adjusted by adding alkalis such as NaOH and KOH.

The demand for electrical throughput continued to increase, thus requiring even higher circuit densities and performance. At present, it is desirable to fabricate a chip having a circuit pattern of eight or more layers. In principle, the need for more layers does not change the nature of the polishing, but this requires a more stringent specification from the polishing method. The width of each layer can be <5 μm. Defects such as scratches and dishing should be reduced or eliminated. Larger wafers make it difficult to maintain uniformity over larger length scales compared to 8 ", or 200 mm wafers.

In addition to adding layers, increased circuit density can be achieved by reducing the space between the individual paths. The paths cannot be too close because electrical spillover can appear across the SiO ₂ dielectric (wafer oxide) to effectively interrupt the connection. Recent technological advances that allow the fabrication of very small high density circuit patterns on integrated circuits place greater demands on isolation structures.

US Patent Application Publication No. 2003/0129838, filed Dec. 28, 1999, describes the following non-plated abrasive materials: iron oxide, strontium titanate, apatite, dioptase. ), Iron, brass, fluorite, hydrated iron oxide, and azurite.

Summary of the Invention

High performance polishing is required in the fabrication of integrated circuits (ICs) for computer and electronics applications. In essence, an IC is a device composed of multiple thin layers deposited sequentially on an inorganic oxide wafer. These layers have different compositions including oxides, metals, or dielectric materials, each of which must be polished within a range of narrow tolerances and high selectivity to obtain a working device. Chemical mechanical polishing (CMP) is the means to accomplish this task. Polishing is accomplished through the removal of surface features using liquid chemical slurries and rotating polymer brushes. In an effective system, the synergistic correlation between surface etch chemicals, surface protective chemicals, abrasives, and polymer pad physics in the slurry creates a uniform flat surface. In the present invention, particles having non-spherical morphology are used as abrasives in CMP slurries.

1 is an example of one modified non-spherical particle of the present invention.

2 is an example of another modified non-spherical particle of the present invention.

3 is an example of a partially coated non-spherical particle of the present invention.

4 is an example of another partially coated non-spherical particle of the present invention.

5 is an example of another partially coated non-spherical particle of the present invention.

6 is an example of another partially coated non-spherical particle of the present invention.

7 is an example of a fully coated non-spherical particle of the present invention.

8 is an example of another partially coated non-spherical particle of the present invention.

9 illustrates the use of CMP to remove riders from the silicon dioxide layer.

10 illustrates polishing of an etched semiconductor wafer.

11 illustrates polishing of an etched wafer containing metal.

12 is a Scanning Electron Micrograph (SEM) of the ultrafine abrasive particles prepared in Example 1 below.

FIG. 13 is a graph comparing the removal rate of copper using a CMP slurry containing aluminum oxide and a CMP slurry containing calcined kaolin particles as an abrasive.

14 is a scanning electron micrograph (SEM) of the ultrafine abrasive particles (Sample A) prepared in Example 3 below.

Generally, CMP slurry compositions contain at least one of abrasives for mechanical action and other chemicals for providing chemical reactions such as oxidizing agents, acids, bases, complexing agents, surfactants, dispersants, and oxidation reactions on the polished surface. Include. Certain poisons are generally avoided. Examples include metal ions with high mobility, such as Na ⁺ , or elements that undergo reaction with the wafer material, such as fluorine (sometimes HF is used in post-CMP cleaning).

Non-limiting examples of available bases include KOH, NH ₄ OH, and R ₄ NOH. Acids, which can be exemplified by H ₃ PO ₄ , CH ₃ COOH, HCl, HF and the like, can also be added. Available as supplementary oxidants by themselves are H ₂ O ₂ , KIO ₃ , HNO ₃ , H ₃ PO ₄ , K ₂ Fe (CN) ₆ , Na ₂ Cr ₂ O ₇ , KOCl, Fe (NO ₃ ) ₂ , NH ₂ OH and DMSO. Divalent acids such as oxalic acid, malonic acid and succinic acid may be used as additives for the polishing composition of the present invention.

Further suitable acid compounds that may be added to the slurry composition include, for example, formic acid, acetic acid, propanoic acid, butanoic acid, pentanic acid, hexanoic acid, heptanoic acid, octanoic acid, nonanoic acid, lactic acid, nitric acid, sulfuric acid, malic acid, tartaric acid. , Gluconic acid, citric acid, phthalic acid, pyrocatechol acid, pyrogallol carboxylic acid, gallic acid, tannic acid, and mixtures thereof.

Suitable corrosion inhibitors that may be added to the slurry composition include, for example, benzotriazole, 6-tolyltriazole, 1- (2,3-dicarboxypropyl) benzotriazole, and mixtures thereof.

If carboxylic acid is added, this can also impart corrosion inhibitive properties to the slurry composition.

Fluorine-containing compounds may be added to the slurry composition to increase the selectivity of the tantalum and tantalum compositions relative to silicon dioxide. Suitable fluorine-containing compounds include, for example, hydrogen fluoride, perfluoric acid, alkali metal fluoride salts, alkaline earth metal fluoride salts, ammonium fluoride, tetramethylammonium fluoride, ammonium bifluoride, ethylenediammonium difluoride, Diethylenetriammonium trifluoride, and mixtures thereof.

Suitable chelating agents that can be added to the slurry composition include, for example, ethylenediaminetetraacetic acid (EDTA), N-hydroxyethylethylenediaminetriacetic acid (NHEDTA), nitrilotriacetic acid (NTA), diethylclentriaminepentaacetic acid ( DPTA), ethanol diglycinate and mixtures thereof. Chelating agents can assist in softening metallic surfaces, or can also help protect the lowered shapes or surfaces of certain compositions. Protective mechanisms can lead to significant improvements.

Suitable amines that may be added to the slurry composition include, for example, hydroxylamine, monoethanolamine, diethanolamine, triethanolamine, diethylene glycol amine, N-hydroxyl ethyl piperazine and mixtures thereof.

Suitable surfactant compounds that can be added to the slurry composition include, for example, any of a number of nonionic, anionic, cationic or amphoteric surfactants known to those skilled in the art.

The pH of the slurry is absolute for the performance of all slurry components. The acidity level of the solution can control the reaction rate on the surface, the formation constant of the metal complexing agent, the rate of the surface oxidation reaction, the solution ionic strength, the aggregation size of the slurry particles, and the like. Examination of various acid, base and pH buffers is an expected area for CMP development.

According to the present invention there is provided a CMP slurry in which the abrasive is formed of particles having at least one dimension (height, length and / or width) having a substantially larger morphology than the other. For the purposes of the present application, this morphology is described as "non-spherical." Thus, the non-spherical particle morphology has at least one dimension that is substantially larger than plate, sheet, needle, capsule, laminar, or the other. It may be in a myriad of other forms. This morphology differs from spherical particles that are substantially round in appearance and do not have a significantly elongated surface. Lamina clays (which may be peeled) such as kaolin, vermiculite and montmorillonite, and variants of these clays that preserve the clay form, such as acid leached kaolin, mica, talc, graphite flakes, glass flakes and synthetic polymer flakes Is useful as abrasive in the CMP slurry of the present invention.

These non-spherical particles are predominant in the slurry. Thus, the phrase "non-spherical particles" as used herein does not include non-spherical aggregation of spherical particles.

Abrasive particles having a non-spherical morphology are believed to offer advantages over ceramic oxide materials of the spherical form in the prior art. It is believed that the pressure of the non-spherical abrasive on the substrate surface is distributed throughout the area, not the contact point as spherical particles. Thus, non-spherical particles provide a gentle polishing action and nevertheless reduce micro-scratching, oxide loss as well as dishing and corrosion as compared to contact point polishing obtained by currently used hard ceramic abrasives.

In addition to having a non-spherical morphology, the abrasive particles are preferably softer than the silica or alumina abrasives generally used for CMP. Thus, the non-spherical abrasive particles have a Mohs hardness of about 1-6. For reference, Table 1 below describes various metal and abrasive particles.

matter Morse Microhardness [㎏ ㎜ ^-2 ] Copper 2.5-3.0 80 tantalum 6.5 230 tungsten 7.5-8.0 350 Hydrated SiO ₂ 4-5 400-500 SiO ₂ 6-7 1200 Copper oxide 3.5-4.0 - Kaolin (function) 2-3 - Kaolin (calcined) 4.0-6.0 Alumina 9.0 2000 ZrO ₂ 6.5 - Diamond 10.0 10000

Non-spherical abrasives with Mohs hardness between about 1-6 are hard enough to provide the required mechanical action of the CMP slurry, but can nevertheless avoid defects such as scratching, dishing and overpolishing action. It is believed to be.

In general, non-spherical particle abrasives may comprise 20% by weight of the slurry, but abrasive solids content of up to 60% by weight may be prepared. More generally, abrasive content in amounts of less than 15% by weight, more preferably in amounts of 0.5-8% by weight, is used.

Currently, kaolin clay particles are preferred as non-spherical abrasives. Although hydrous kaolin can be used, it has been found that when kaolin is calcined a better polishing rate is obtained. However, the overall performance of hydrous kaolin is better than calcined kaolin, so hydrous kaolin is preferred. Calcination of kaolin undergoing a strong endothermic reaction associated with dehydroxylation results in metakaolin. Kaolin clay calcined under more intense conditions than those used to convert kaolin to metakaolin, i.e. kaolin clay calcined to undergo a characteristic kaolin exothermic reaction, is a spinel form of calcined kaolin, and also more extreme If the condition is used, it creates a mullite. In general, calcination of hydrous kaolin at 1200 ° F. and higher temperatures results in dehydroxylation of hydrous kaolin to metakaolin. A calcining temperature of 1400-2200 ° F. can be used to produce kaolin clay calcined with spinel-type kaolin through its characteristic exothermic reaction. At higher temperatures, for example above 1900 ° F., the formation of mullite occurs. All of these forms of kaolin clay can utilize commercially available products from Engelhard Corporation, Iselin, New Jersey, the assignee of the present invention.

The hydrous kaolin is generally prepared through a combination of unit operations that modify the particle size distribution and remove coloring impurities from the kaolin. These unit operations are facilitated by using an aqueous suspension of kaolin in water. Examples of unit operations that change the particle size distribution are centrifugation, delamination or grinding apparatus and selective flocculation. Examples of unit operations causing removal of coloring impurities are suspension and magnetic separation. In addition, reductive and / or oxidative bleaching can be used to render the coloring impurities colorless. Filtration can also be used to substantially remove water from the kaolin, followed by spray drying the high solids content of the filtration product slurry. The spray dried portion may be added back to the high solids content filtration product slurry to further raise the solids content of the slurry. The filtration product cannot be dispersed and thus the filtercake can be dried and powdered to obtain what is called industrially acid dried kaolin product. In addition, kaolin can be modified by thermal or chemical treatment. Generally, kaolin is powdered before and after the calcination operation. Treated kaolin may also be slurried to further modify the particle size distribution through the above-mentioned operation.

Other useful non-spherical abrasive particles are brucite (magnesium hydroxide), hydrotalcite, and nanotalk. The above-mentioned materials can use a commercial item. Other useful non-spherical abrasive particles are described in US Pat. No. 6,187,710, commonly assigned as incorporated herein by reference. In one embodiment, this patent is a basic three-layer consisting of a central layer (octahedral layer) of metal ions surrounded by oxygen in octahedron and a silicon atom-containing layer (tetrahedral layer) surrounded by two tetrahedrons surrounding the layer. Clay minerals, which consist of platelets and vary in dimension of clay particles from 0.1 micron to 1 micron. Up to 30 at.% Of metal ions are replaced by lower valence ions in the octahedral layer and up to 15 at.% Of silicon ions are replaced by lower valence ions in the tetrahedral layer. This patent suggests that in another embodiment, silicon (germanium) can be replaced by trivalent ions in the tetrahedral layer. In the octahedral layer preferably aluminum, chromium, iron (III), cobalt (III), manganese (III), gallium, vanadium, molybdenum, tungsten, indium, rhodium, and / or scandium are present as trivalent ions. As divalent ions, magnesium, zinc, nickel, cobalt (II), iron (II), manganese (II) and / or beryllium are preferably present in the octahedral layer. In the tetrahedral layer, silicon and / or germanium are present as tetravalent components, preferably aluminum, boron, gallium, chromium, iron (III), cobalt (III) and / or manganese (III) as trivalent components. do.

The necessary ingredients for the synthesis, oxides of silicon (germanium) for the tetrahedral layer and 3/2/1 valent ions for the octahedral layer are placed in an aqueous medium and the desired pH (3-9, preferably 5- 9), then hold at a temperature of 60-350 ° C. for a predetermined time and keep the pH within the desired range. The reaction time depends greatly on the temperature and thus the pressure, and at higher temperatures a shorter reaction time is possible. In practice, reaction times on the order of 5-25 hours are found at lower temperatures of 60-125 ° C., while reaction times of a few minutes to about 2.5 hours may be sufficient at temperatures of 150 ° C. and above. The reaction time determines the dimensions of the clay mineral in part.

This method can be carried out in various ways depending on the nature of the components and the desired result. Preferably, the chlorides of the metals involved do not work together because they induce a reaction to clay minerals, which are not at all but are hardly detectable. For further details on the method, refer to US Pat. No. 6,187,710, incorporated.

Another useful non-spherical abrasive particle includes expandable clay platelets modified through complexation with other components. Expandable clay systems include smectite clay, montmorillonite, laponite, stevensite, and many other natural and synthetic clays with various compositions, charge densities, and platelet dimensions. In the clay literature it is known that these types of layered materials can be modified by various ion exchange and intercalation methods. In addition, positively charged platelets, such as hydrotalcite and other layered double hydroxides, can undergo a similar type of chemistry to negatively charged smectite platelets.

One example of an expandable clay platelet modified through complexation with other components is as follows. In FIG. 1, the expandable clay platelets 10 have charged cations 12 such as sodium ions that reside in the interlayer space of the clay platelets. Expandable clay platelets 10 are ion exchanged with inorganic clusters 14 such as aluminum oxide hydroxide ions (Al ₁₃ Keggin ions) to displace cations 12. The higher charge densities of these resulting clusters provide stronger interlayer interactions and the clay layer remains stacked. The resulting material is used without further deformation or heated to elevated temperatures to form a three-dimensional columnar structure. Alternatively, an anionic cluster such as poly-oxometallate of Mo, W and other transition metals may be inserted in the platelet charged in the same amount as hydrotalcite.

Another example of expandable clay platelets modified through complexation with other components is as follows. In FIG. 2, the expandable clay platelets 10 have charged cations 12 such as sodium ions that reside in the interlayer space of the clay platelets. Organic cations 16 such as long chain alkyl ammonium ions are exchanged into the interlayer space of smectite or similar negatively charged expandable clay platelets 10. Alternatively, anionic organic molecules, such as organo-sulfonates, may be inserted into the interlaminar space of platelets charged in the same amount as hydrotalcite.

Another useful non-spherical abrasive particle includes a central host coated with a secondary component. The central host may be non-spherical particles such as those described above, three-dimensional particles such as alumina, or other metal oxide particles. The coating can partially or completely cover the central host. In addition, the particles may have multiple coatings thereon.

One example of a partially coated platelet is as follows. In FIG. 3, host non-spherical particles 18 are partially coated by smaller platelets 20. Examples of useful smaller platelets 20 include organic polymers coated on the surface of laponite or other smectite particles or host non-spherical particles 18. Certain smaller platelets may have the desired degree of softening and composition, but platelet sizes may be too small, or may have problems in dispersing them. By placing these platelets on the surface of the host, more effective abrasives with more desirable rheology, dispersion, and the like can be developed. These types of materials can be synthesized by a number of approaches, including layer-by-layer techniques.

Another example of a partially coated platelet is as follows. In FIG. 4, host non-spherical particles 18 are coated by substantially smaller spherical particles 22. Examples of useful smaller spherical particles 22 include colloidal particles such as colloidal silica and molecular species such as aluminum oxide hydroxide ions. The size, composition, charge density and other properties of the smaller spherical particles 22 can be adjusted to suit the desired final properties. Subsequent coatings may also place spheres of different sizes on the surface to produce different levels of packing, porosity and softness.

Another example of a partially coated platelet is as follows. In FIG. 5, host non-spherical particles 18 are coated by substantially smaller microcrystals 24. Examples of useful smaller microcrystals 24 include metal oxide or silica microcrystals or non-oxide ceramic phases such as metal carbides and nitrides. Such coatings may be formed by converting platelets or colloidal particles into crystalline oxides by heating. Alternatively, the desired phase may be crystallized directly on the surface of the host non-spherical particle 18 similar to known techniques for forming titanium dioxide coated mica pearlescent pigments. Examples of useful methods are described in US Pat. No. 4,038,099, commonly assigned as is incorporated herein by reference.

Another example of a partially coated platelet is as follows. In FIG. 6, host non-spherical particles 18 are substantially coated by polymer 26. Examples of useful polymers include diallyldimethylammonium chloride (abbreviated PDADMAC) or polysodium styrene sulfonate (abbreviated PSS). Surface properties such as charge, softness, isoelectric point, rheology and the like can be adjusted by coating the surface of host non-spherical particle 18 with polymer 26.

Examples of fully coated platelets are as follows. In FIG. 7, host non-spherical particles 18 are completely coated by carbon 28. Various precursors such as polymers, organic molecules, and the like may be disposed on the surface of the host particle 18. Thereafter, the coated material is pyrolyzed to form carbon coating 28. The carbon coating can be very thin (thickness several millimeters) or thick, depending on the desired properties.

Another example of a partially coated platelet is as follows. In FIG. 8, host non-spherical particle 18 is partially coated by organic functional group 30. Particle 18 may be treated with a coupling agent such as an organo-silane to attach the molecule directly to the particle surface. Generally, reactive groups on the particle surface, such as hydroxide groups, react with the alkoxy or halo groups of the silanes. The result is the introduction of organic groups with specific functionality to the particle surface.

Regardless of the type used, the particle size of the non-spherical abrasive may be measured by commercially available particle measurement techniques and may generally have an average diameter of less than about 1 micron. See, for example, commonly assigned U.S. Pat.No. 4,767,466, which measured particle size by a Sedigraph 5100 particle size analyzer and reported the corresponding spherical diameter based on weight percentage. . For example, kaolin particle size is measured by x-ray sedimentation such as Cedigraph 5100. The average particle size for kaolin may preferably range from about 0.01 to less than about 1 micron, more preferably from about 0.01 to about 0.5 micron.

Non-spherical abrasives may be combined with any chemical adjuvant that generally forms CMP slurries, such as acids, bases, dispersants, oxidizing agents, complexing agents, surfactants, and / or passivating agents. CMP slurries containing non-spherical abrasives can be used in the CMP process. Examples of representative CMP processes are described below. These are presented by way of example only and are not provided for the purpose of limiting the use of the CMP slurries of the invention to the particular process techniques or conditions described. Accordingly, CMP slurries of the present invention containing non-spherical abrasives are intended to be used in any CMP method that is currently known or may be used in the future as the complexity of integrated circuits increases.

For example, in oxide-CMP the pH of the aqueous solution is maintained to soften the surface of the silicon wafer so that the suspension of small particles is maintained and the high shape can be crushed and removed by the action of the abrasive. Depending on the chemistry chosen, the pH of the slurry can be adjusted accordingly. Thus, the pH can be acidic or basic. The surface of the wafer is believed to cause deformation under alkaline conditions as schematically shown in FIG. As shown in Fig. 9, the substrate 32 formed of silicon dioxide is treated by a combination of chemical (alkaline reactivity) and mechanical action (particle polishing). This situation represents the simplest case of oxide-only polishing. Thus, the silicon-oxide-silicon bonds are broken by alkaline reaction, and individual silicon hydroxide sites on the surface are removed by mechanical polishing.

To place the electrical circuit on the chip, the pattern must be etched on the wafer surface as shown in FIG. In this embodiment, the substrate 34 is etched to form a series of channels 36 that can be filled with a dielectric or conductive metal component. The etched substrate 34 increases the attack of polishing because the surface is not uniform. Substrate 34 as shown has a low patterned etched region (A) and a high patterned density (B). Since the local pressure exhibited by the pad is distributed over a smaller surface area, surface removal during polishing tends to be greater in the region (B) with a higher pattern density. Other defects such as sharp corners of the pattern and corrosion and rounding of the shape should also be minimized.

Once the wafer containing the etched pattern is fabricated, a metal layer, which may be an electrical circuit, may be applied. 11 illustrates such a wafer comprising a wafer substrate 38, a patterned region or channel 40, and a metal or metal alloy 42 contained within the patterned region. The metal used is generally a conductive copper / aluminum (Cu / Al) alloy or tungsten (W) which is more resistant to temperature and oxidation than bulk Cu metals. Polishing is necessary to eliminate metal overburden 44 as the metal layer extends beyond the low etched region. As opposed to oxide polishing, metal polishing is performed using an oxidant in aqueous solution to form a layer of soft oxides on the metal surface that can be removed by mechanical polishing in the slurry. Again, both chemical and mechanical means are used to polish the surface.

There is an added attack by metal-CMP. Many surface compositions appear with varying coating densities, but uniform removal of metal must be obtained. To prevent electrical shorts between circuit lines, all overloaded metals must be removed. A portion of metal surface 42 may undergo metal overpolishing in trench region 40, referred to as dishing in FIG. 11. In FIG. 11, LS stands for line space, while LW stands for line width. The sum of LW and LS is the pitch. LW divided by pitch is the pattern density. A method for limiting dishing is to add a complexing agent that binds to the low lying metal region to form a protective layer and further limit metal corrosion due to slurry oxidants. Obviously, the aqueous slurry and pad composition should be carefully chosen so that the corrosion and protection methods are balanced.

Removing excess metal or other contaminants from the smaller spaces between the individual paths represents an ever increasing attack for the CMP process. Copper metals have a smaller resistivity and capacitance compared to the Cu / Al alloys currently used as conductive media. Thus, sending a signal through the copper line requires a smaller potential to reduce the tendency of electrical spillover. In fact, by using Cu alone, the circuit paths can be placed closer together.

However, the use of Cu also has disadvantages. Copper does not adhere well to oxide surfaces. Unlike WO ₃ or Al ₂ O ₃ , copper can also be sensitive to bulk oxidation because CuO or CuO ₂ surface layers still allow O ₂ and H ₂ O to penetrate into the bulk metal. In addition, Cu atoms are mobile and can migrate into the SiO ₂ wafer material, ultimately failing to generate transistors in the circuit. Thus, a thin layer of low dielectric material, generally comprised of tantalum, tantalum nitride or titanium nitride, is disposed between the wafer oxide and the conductive Cu layer. The buffer layer promotes Cu adhesion, prevents oxidation of the bulk Cu metal, prevents Cu ion contamination of the bulk oxide, and also lowers the dielectricity between the circuits (ie, the circuits get closer together).

One of the uses of CMP technology is the fabrication of shallow trench isolation (STI) structures in integrated circuits formed on wafers such as semiconductor chips or silicon. The purpose of the STI structure is to isolate discrete device elements (eg, transistors) in a given pattern layer to prevent current leakage between them.

STI structures are generally formed by thermally growing an oxide layer on a silicon substrate and then depositing a silicon nitride layer on the thermally grown oxide layer. After depositing the silicon nitride layer, the shallow trench is, for example, through any of the well-known photolithography shielding and corrosion processes through the silicon nitride layer and the thermally grown oxide layer, and in part through the silicon substrate. Is formed. Thereafter, a layer of dielectric material, such as silicon dioxide, is generally deposited using chemical vapor deposition to completely fill the trench and cover the silicon nitride layer. The CMP process is then used to remove that portion of the silicon dioxide layer covering the silicon nitride layer and to planarize the entire surface of the article. The silicon nitride layer seeks to act as a polishing stop that prevents the underlying thermally grown oxide layer and silicon substrate from being exposed during the CMP process. In some applications, the silicon nitride layer is later removed, for example by immersing the product in HF acid solution, such that only silicon dioxide filled trenches act as STI structures. Thereafter, further processing is generally performed to form a polysilicon gate structure.

The use of Cu and the accompanying low dielectric buffer layer requires enhanced efficacy from the polishing technique. The new technique is called Cu-CMP, but in principle does not differ significantly from previous polishing methods. The CMP method must remove soft Cu metal overload, but should be able to limit Cu dishing, scratching and low dielectric buffer layers. At the same time, the tolerance is tighter due to the more closely spaced circuit patterns. The ability to produce thin, flat and flawless layers is of utmost importance.

As is also known in the art, one method of forming internal interconnects in semiconductor structures is the so-called dual damascene process. The dual damascene method begins with the deposition of a dielectric layer, typically an oxide layer, disposed on a circuit formed of single crystals, for example silicon. The oxide layer is corroded to form a trench having a pattern corresponding to the pattern of vias and wires for internal wiring of elements of the circuit. Vias are openings in oxide through which different layers of the structure are electrically interconnected, and the pattern of wires is defined by trenches in the oxide. Thereafter, metal is deposited to fill the openings in the oxide layer. Subsequently, excess metal is removed by polishing. This process is repeated as many times as necessary to form the necessary internal wiring. Thus, the dual damascene structure has a trench in the upper portion of the dielectric layer and a via that terminates at the bottom of the trench and passes through the lower portion of the dielectric layer. This structure has a step between the bottom of the trench and the sidewalls of the vias at the bottom of the trench.

The abrasive particles of the present invention may be used in applications other than logic (e.g., microprocessors) or memory (e.g. flash memory) devices in which copper is used in the interconnecting metallic layer. Can be used for CMP. For example, improving the thermal and electrical characteristics of the packaging of the device may include the use of a copper layer that needs to be planarized. The structure of the interconnect copper layer in the integrated circuit device and the copper layer in the packaging can be different, causing different requirements for the thickness, flatness, dishing and defects of the layer to be removed. In addition, Micro-ElectroMechanical Systems (MEMS) may have a copper layer that may require planarization using CMP. The abrasive particles of the present invention can also be used in these applications in CMP slurries.

A review of the CMP process is presented in "Advances in Chemical-Mechanical Planarization," Rajiv K. Singh and Rajiv Bajaj, MRS Bulletin, October 2002, pages 743-747. In general, the CMP method appears to be very simple, but gaining a detailed understanding is mainly limited by the large number of input variables in the polishing method. Among these variables are slurry variables such as particles and chemicals, pad variables, equipment variables such as down pressure and linear velocity, and substrate variables such as pattern density. This paper provides a good review of new applications for process variables and CMP technology and is incorporated herein by reference.

Example  One

Ansilex 93 ™ calcined kaolin slurry (50% solids) supplied by Engelhard Corporation was used as starting material. The slurry was mixed with 4 pounds / ton of Defloc 411 (ammonium polyacrylate) supplied by Sharp Specialty Chemicals. This mixture was ground using Netzsch with zirconia beads at 1.2 gpm (gallons per minute)-2 passes. Two pounds of Deflock 411 was added again per tonne after the Netssh mill, and the mixture was spray dried to prevent damage to the slurry. The spray dried product was reslurried in Waring Blender for 5 minutes and then deslimed to 40% solids without restriction for 26 minutes on CU5000 (centrifuge). Ultra-viscosity of ultrafine fractions of interesting particulate slurries was taken by CMP applications. The size distributions of the spray dried and ultrafine products as measured by Cediggraph 5100 are described in Table 2 below. The SEM of the ultrafine product diluted several times to enhance the image quality is shown in FIG. 12. SEM was obtained using a long emission electron microscope (Jol 6500F) at 5 kV.

Annecyx 93 Slurry Reslurried Spray Dried Product Ultrafine Product PSD (finer mass%) (microns) 210.50.30.2 927946165 9483583117 10099988565

Example  2

The ultrafine product of Example 1 was reslurried to 4% solids. The slurry was passed through a Puradisc 25 GD glass filter (25 mm diameter and pore size of 2 microns) to remove too large particles. Chemical packages from common copper CMP slurries were added to the polishing slurry. Chemical packages included oxidizing agents (hydrogen peroxide), passivating agents (benzotriazoles), complexing agents / etching agents (citric acid) and stabilizers (TEA, TX-100). For comparison, a commercially available alumina-based CMP slurry (Cabot Microelectronics) was used.

CMP slurries were tested on 200 mm Si wafers provided with copper interconnects and Ta diffusion barriers by the double-damascene method. The CMP slurry was applied using a polishing machine (Novellus IPEC 372) with a down pressure of 2 psi. The results are shown in Table 3 and FIG. In addition, reference is made to FIG. 11 to understand the definitions of "pitch" and "pattern density" used in Table 3.

Surface Morphology (Overpolished Area)  Alumina slurry  Kaolin slurry  100 μm pitch  100 μm pitch  50% pattern density  50% pattern density  Corrosion and dishing-violent overpolishing  Corrosion <1 nm  Very low Cu / Ta selectivity  Excellent Cu / Ta Selectivity  1. Pitch = Line Width + Line Space 2. Pattern Density = Line Width / Pitch

The measurements in Table 3 are made from overpolished regions of the wafer. In the case of alumina slurries, vigorous overpolishing hindered measurements for dishing and corrosion.

Example  3

A hydrous kaolin spray dried product provided from Engelhard was used as starting material. The spray dried product was reslurried at 40% in a Waring blender for 5 minutes in the laboratory and then viscous to 40% solids without limitation (2400 rpms) for 15 minutes on a CU5000 centrifuge. The ultrafine hydrous kaolin fraction, which constitutes the supernatant of 5% solids from the deviscosation step, was filtered through Whatman filters (25 mm diameter and pore size 2 μ) to make up the polishing slurry for use in CMP formulations. (Sample A). The size distributions of the starting spray dried product and the ultrafine product as measured by Cediggraph 5100 are described in Table 4.

PSD (finer mass%) (microns) Starting spray dried product Ultrafine function kaolin (sample A) 210.50.30.18 9898927350 10099988461

The SEM of the ultrafine hydrous kaolin (Sample A) diluted several times to enhance the image quality is shown in FIG. 12. SEM was obtained using a long emission electron microscope (Jeol 6500F) at 10 kV.

Another sample (Sample B) was prepared from a different starting feed PSD material and the ultrafines were further 32 at 2400 rpm after viscous with 40% solids without restriction (2400 rpms) for 15 minutes on a CU5000 centrifuge. It was further applied to dedusting and filtered through Whatman filter (25 mm diameter and pore size 2 μ). The size distributions of the starting spray dried product and the ultrafine product as measured by Cediggraph 5100 are described in Table 5.

PSD (finer mass%) (microns) Starting spray dried product Ultrafine function kaolin (sample B) 210.50.30.18 8880664732 10099999378

Example  4 and Comparative example  A and B

The chemical package from Example 2 was added to the ultrafine hydrous kaolin slurry from Example 3 (Sample A) to prepare a CMP formulation for planarization of Cu (Example 4). Other CMP formulations were prepared using the same chemical package using fumed silica slurries and alumina slurries. The fumed silica used was Aerosil 200 provided by Degussa (injector size of 12 mm and average aggregate size of 170 mm as measured by Microtrac) (Comparative Example A ). The alumina particles were in alpha form and were obtained from Polishing Solutions Inc. (Comparative Example B). Commercially available alumina particles are used in commercial CMP slurries for metal planarization.

CMP slurries are tested on uncoated 200 mm tetraethylorthosilicate (hereinafter referred to as "TEOS") silica wafers and wafers coated with copper or tantalum to predict polishing time for removing copper on patterned wafers. Not only was the polishing rate measured for help, but also the surface smoothness and selectivity between copper / tantalum and copper / silica. The CMP slurry was then tested on a 200 mm Si wafer (patterned wafer) provided with a copper interwire and a Ta diffusion barrier by the dual damascene method to evaluate corrosion and dishing. Corrosion was measured at 70% patterning density while dishing was measured on a 300 micron pitch copper line. Dicing and corrosion measurements were performed on both polished and overpolished wafers (20% extended time overpolished wafers) to determine the selectivity for overpolishing of these undesirable morphological shapes.

The test was performed on an IPEC-372 M machine (Polishing Solutions Inc., Phoenix, AZ) at 5 psi down pressure and 60 rpm platen speed. WIWNU means With Wafer Non Uniformity.

Blanket  Wafer ( Blanket Wafer )

Abrasives in CMP Slurry Material removal rate (mm / min) WIWNU% Cu / Ta selectivity Tantalum / TEOS Selectivity Example 2 254 4.7 65 5.9 Comparative Example A 206 4.9 40 1.4 Comparative Example B 713 10.2 17 2.8

Clearly, ultrafine hydrous kaolin based CMP slurries provided the desired high selectivity and uniformity over fumed silica or alumina. The removal rate of copper material by the ultra fine hydrous kaolin is equivalent to that of fumed silica and lower than that by alumina. Cu / Ta selectivity is more critical than polishing rate because the expected results from the Cu planarization slurry stop in the Ta layer. Low Ta planarization rates with hydrous kaolin formulations make it impossible to take advantage of better Ta / TEOS selectivity over silica or alumina based CMP formulations.

Patterned wafer

Just-polished: Estimated by visual inspection of the wafer when copper is properly removed.

Overpolished: Newly patterned wafer polished over 20% extension time than needed to remove copper

Abrasives in CMP Slurry Polished Condition Overpolished Condition Dishing (nm) Corrosion (nm) Dishing (nm) Corrosion (nm) Example 2 419 95 618 74 Comparative Example A 450 278 655 386 Comparative Example B 332 186 522 390

Abrasives in CMP Slurry % Increase in dishing Percent increase in corrosion Example 2 50 0 Comparative Example A 50 40 Comparative Example B 60 110

Ultrafine hydrous kaolin based CMP slurries provided significantly lower corrosion without selectivity for overpolishing compared to silica and alumina. This is consistent with high selectivity for copper / tantalum and tantalum / TEOS removal rates obtained using ultrafine hydrous kaolin slurries. Thus, ultrafine hydrous kaolin slurries are expected to provide lower corrosion as well as oxide and metal losses as compared to silica and alumina.

The dishing was similar to all abrasives, suggesting an important role of the chemical in the formulation relative to the mechanical action of the abrasive.

Example  5 and Comparative example  C and D

In this example, the ultrafine hydrous kaolin (Sample B) and fumed silica in Example 4 except removing TX100 and TEA from the chemical package and lowering the pH of the slurry to 5-4 (Comparative Example C) A CMP formulation based on was used. In addition, an alumina-based commercially available slurry (CCMP) provided by Cabot Microelectronics was also used (Comparative Example D).

CMP slurries were tested on uncoated 200 mm TEOS wafers and wafers coated with copper or tantalum to aid in calculating polishing times for patterned wafers, as well as copper / tantal as well as copper / tantal as well as polishing rates, surface smoothness and copper / tantalum. Selectivity between silicas was measured. The CMP slurry was then tested on a 200 mm Si wafer (patterned wafer) provided with a copper interwire and a Ta diffusion barrier by the dual damascene method to evaluate corrosion and dishing. Corrosion was measured at 70% patterning density while dishing was measured on a 150 micron wide copper line. Dicing and corrosion measurements were performed on both polished and overpolished wafers (20% extended time overpolished wafers) to determine selectivity for overpolishing.

The test was performed at the down pressure of 2 psi and the platen speed of 90 rpm on the same machine as in Example 4.

Blanket  wafer

Abrasives in CMP Slurry Material removal rate (mm / min) WIWNU,% Cu / TEOS selectivity Example 5 173 One 1020 Comparative Example C 224 One 83 Comparative Example D 127 17 52

Clearly, ultrafine hydrous kaolin based CMP slurries provided improved material removal rates and significant improvements in selectivity and uniformity over commercial slurries. Although the removal rate is slightly lower than that of silica, this factor is much more weighted by the dramatic increase in Cu / TEOS selectivity. The very high selectivity is due to the extremely low removal rate for TEOS by the ultra fine water, which provides low corrosion and which oxide circuit designers typically have to compensate for at design time, and thus little metal loss. .

Patterned wafer

Properly-polished: Estimated by visual inspection of the wafer when copper is properly removed.

Overpolished: Newly patterned wafer polished over 20% extension time than needed to remove copper

Abrasives in CMP Slurry Polished Condition Overpolished Condition Dishing (nm) Corrosion (nm) Dishing (nm) Corrosion (nm) Example 5 405 25 510 32 Comparative Example C 503 223 466 247 Comparative Example D 131 45 133 70

The ultrafine hydrous kaolin based CMP slurry provided approximately 10% of corrosion due to silica and 50% of corrosion due to commercial CCMP slurries. Commercial slurries showed better dishing compared to silica and hydrous kaolin samples, demonstrating the importance of optimized chemicals in the formulation.

This patent application prioritizes pending US patent application Ser. No. 60 / 455,216, filed Mar. 17, 2003, which is incorporated herein in its entirety, and Ser. No. 60 / 509,445, filed Oct. 8, 2003. .

Claims (10)

A chemical-mechanical planarization polishing slurry comprising primary abrasive particles having a non-spherical morphology. The slurry of claim 1 wherein the abrasive particles having a non-spherical morphology are selected from mica, talc, lamina clay, graphite flakes, glass flakes, and synthetic polymer flakes. 3. The slurry of claim 2 wherein the abrasive particles comprise lamina clay. The slurry of claim 3, wherein the clay particles are calcined at a temperature of at least 1200 ° F. 5. The slurry of claim 1 comprising up to about 20 wt% abrasive particles. 6. The slurry of claim 5 comprising from about 0.5 to about 8 weight percent abrasive particles. The slurry of claim 1, wherein the abrasive particles have an average diameter of less than about 1 micron. 8. The slurry of claim 7, wherein the abrasive particles have an average diameter of about 0.01 to about 0.5 microns. The slurry of claim 1, wherein the abrasive particles have a Mohs hardness in the range of about 1 to about 6. 6. The slurry of claim 1, wherein the non-spherical abrasive particles are modified or at least partially coated.