TW202523818A - Compositions and methods for removal of tungsten and dielectric layers - Google Patents

Abstract

The presently claimed invention relates to dielectric polishing composition and methods thereof. The presently claimed invention particularly relates to a composition comprising: (A) surface - modified colloidal silica particles comprising a negatively - charged group on the surface of the particles, wherein the surface-modified colloidal silica particles have a negative charge, a particle size of from 60 nm to 200 nm, and a zeta potential ≤ -35 mV at a pH in the range of from ≥ 2.0 to ≤ 6.0; (B) at least one corrosion inhibitor selected from at least one guanidine derivative; (C) at least one iron (III) oxidizer; (D) at least one buffering agent selected from at least one acidic amino acid having an isoelectric point (pI) of ≤ 5.0; (E) at least one stabilizer; and (F) an aqueous medium, wherein the pH of the composition is in the range of from ≥ 2.0 to ≤ 4.3, and wherein the composition has a polyacrylamide content ≤ 1 ppm.

Description

Compositions and methods for removing tungsten and dielectric layers

The claimed invention relates to compositions and methods for polishing dielectrics. In particular, the claimed invention relates to compositions and methods that provide high selectivity for removing dielectrics relative to the tendency of tungsten to do so.

The manufacturing of integrated circuits (ICs) is a multi-step process. The first step (also called the front-end of the line or FEOL step) consists of patterning individual components, such as transistors (e.g. C-MOSFET) on the semiconductor by suitable techniques (such as photolithography or ion implantation). This is followed by the insertion of metal wires/pillars and insulating layers (such as dielectrics) to interconnect the various components (also called the back-end of the line or BEOL step). As the feature size in ultra-large scale integrated circuit (ULSI) technology continues to shrink, multiple layers of interconnects must be formed.

Metals such as tungsten, cobalt, ruthenium, and copper are often used for connections and the choice of a specific metal is done based on its location in the architecture. For example, typically the first metal interconnect (metal level 0) is a tungsten pillar inserted into a dielectric layer such as silicon dioxide. On the other hand, during the BEOL steps, copper is the most commonly deposited metal, but for lower interconnect stages (e.g., metal levels 1 – 4), tungsten and cobalt are also commonly utilized.

CMP (Chemical Mechanical Planarization) has been found to be a key enabling technology for multi-level interconnect formation, as it can induce both local and global planarization and at the same time provide good surface quality in terms of an excellent mirror-like defect-free surface finish. Pure mechanical polishing or (fine) grinding provides excellent planarization but a matte surface. On the other hand, pure chemical polishing (anisotropic etching) provides only poor planarization results but an excellent mirror-like surface finish.

In this context, CMP utilizes the interaction of chemical and mechanical actions to achieve the desired planarity and finish of the surface to be polished. The mechanical action is usually achieved by the interaction of a CMP composition, which includes a finely dispersed abrasive, with a polishing pad, which is typically pressed onto the surface to be polished and mounted on a moving platform. In a typical CMP process step, a wafer holder is rotated to bring the wafer to be polished into contact with the polishing pad. The CMP composition is usually applied between the wafer to be polished and the polishing pad.

A typical CMP composition or slurry contains one or more abrasive (insoluble/dispersed) components plus a soluble component. The chemical action is provided by the soluble components of the CMP composition. Generally, to achieve metal CMP, the chemical compositions are usually tailored by balancing corrosive components (such as acids, bases or oxidizers) with suitable inhibitors. However, it is important that the soluble and insoluble components should be compatible to ensure colloid stability. Solutions or agglomerates that are unstable in terms of colloids can damage the surface and delicate structures on the wafer to be polished due to scratches. Therefore, in order to achieve suitable planarization and surface quality, the components and their concentrations must be accurately selected to obtain a suitable CMP composition.

Additionally, depending on the integration scheme, the characteristics of some or all of the above mentioned metals may have to be considered when designing the slurry (e.g., etching). For example, when fabricating an IC that includes tungsten (metal 0), CMP is used to remove metal and pad/barrier overburden until a planar metal 0 (or higher) layer is exposed.

For example, U.S. 6,083,419 A describes a CMP composition comprising a compound capable of etching tungsten, at least one tungsten etching inhibitor, wherein the tungsten etching inhibitor is a compound comprising at least one functional group selected from the following: a nitrogen-containing heterocyclic ring without a nitrogen-hydrogen bond, a sulfide, an oxazolidine, or a mixture of functional groups in one compound.

In cases such as U.S. 6,083,419 A, since the intended endpoint for such applications is a dielectric (e.g., a silicon oxide layer obtained by CVD deposition using tetraethyl orthosilicate or TEOS), state-of-the-art slurries are tailored to achieve high tungsten versus silicon oxide removal rates. For example, US 2019/0211227 and US 2019/0211228 aim to achieve high selectivity for tungsten removal by utilizing surface-modified colloidal silicon dioxide particles with a negative charge. As is well known, one consequence of using a composition that exhibits high selectivity towards metal (e.g. tungsten) removal is that embedded metal lines/pillars (surrounded on both sides by dielectric) are preferentially corroded to produce dish-like recesses, leading to the name - dishing. However, dishing is not limited to metal vs. dielectric, but can also affect any interface on the surface of a substrate where selectivity of one component vs. the other in terms of removal rate exists. A particularly effective way to combat this is to fully balance the removal rates of the two components across the interface.

In addition, in order to realize new integration schemes, selective removal of dielectrics is highly desired. US 8,492,277 provides a polishing method comprising using a composition comprising a non-cyclic organic sulfonic acid to provide selective high removal rates for silicon oxide and silicon nitride. US 8,513,126 achieves similar results using a composition comprising a quaternary ammonium salt. More recently, WO 2023/186762 A1 discloses a composition that achieves such high selectivity by the presence of a combination of a guanidine derivative, an iron (III) salt, a potassium phosphate salt, and polyacrylamide plus a suitable stabilizer (such as EDTA).

US 2022/0033682 A1 discloses a chemical mechanical polishing composition for polishing tungsten or molybdenum, which comprises, consists essentially of, or consists of a water-based liquid carrier, abrasive particles dispersed in the liquid carrier, an amino acid selected from the group consisting of arginine, histidine, cysteine, lysine, and mixtures thereof, an anionic polymer or anionic surfactant, and optionally an amino acid surfactant.

US 2019/0051537 A1 discloses a polishing composition containing the following as initial components: water; an oxidizing agent; arginine or a salt thereof; a dicarboxylic acid, an iron ion source; a colloidal silica abrasive; and an optional pH adjuster; and an optional surfactant; and an optional biocidal remover.

However, as is well known in the field of semiconductor development, the presence of trace amounts of undesirable impurities such as alkaline metals (e.g., potassium) on the wafer surface can have adverse effects on performance. Therefore, the components utilized in the CMP composition must be carefully selected due to their economic impact and the undesirable footprint they leave on the wafer surface in the form of debris/residues. Thus, there is always room for the development of improved and more economical compositions. Furthermore, in the presence of metals such as tungsten and cobalt that are easily removed in acidic conditions, there is an unmet need to obtain compositions and methods that allow for the removal of dielectrics with high selectivity relative to tungsten and/or cobalt while maintaining high surface quality.

Surprisingly, it has been discovered that the compositions of the presently claimed invention described herein below provide unexpectedly high selectivity for removal of dielectrics over tungsten-preferential removal.

Accordingly, in one aspect of the claimed invention, a dielectric polishing composition comprises: (A) surface-modified colloidal silica particles, the particles comprising negatively charged groups on the surface of the particles, wherein the surface-modified colloidal silica particles have a negative charge, a particle size from 60 nm to 200 nm, and a zeta potential of ≤ -35 mV at a pH ranging from ≥ 2.0 to ≤ 6.0; (B) at least one corrosion inhibitor selected from at least one guanidine derivative; (C) at least one iron (III) oxidizing agent; (D) at least one buffer selected from at least one basic amino acid having an isoelectric point (pI) ≥ 6.5; (E)   at least one stabilizer; and (F)   an aqueous medium, wherein the pH of the composition ranges from ≥ 2.0 to ≤ 4.3, and wherein the polyacrylamide content of the composition is ≤ 1 ppm.

On the other hand, the claimed invention is directed to a process for manufacturing a semiconductor device comprising chemically mechanically polishing a substrate (S) used in the semiconductor industry in the presence of a composition described herein, the substrate (S) comprising: (i)   tungsten and/or (ii)   tungsten alloy; and (iii)  at least one dielectric layer selected from silicon, silicon oxide, silicon nitride or a low-k material.

In another aspect, the claimed invention is directed to the use of the compositions described herein for polishing a substrate (S) comprising: (i) tungsten and/or (ii) a tungsten alloy; and (iii) at least one dielectric layer.

The claimed inventions relate to at least one of the following objects: (1) The claimed compositions and methods are directed to providing selective removal of dielectric layers relative to other metals such as tungsten and/or cobalt. (2) The claimed compositions and methods are directed to providing high silicon oxide (SiO ₂ ) removal rates while ensuring low tungsten (W) removal rates. (3) The claimed compositions and methods are directed to preventing undesirable etching of tungsten (W) during chemical mechanical polishing of tungsten-containing substrates. (4) The claimed compositions and methods are directed to preventing/suppressing undesirable recessing of tungsten during chemical mechanical polishing while maintaining optimal control of dielectric/oxide/TEOS recessing. (5) The objective of the composition of the invention claimed herein is to provide a stable formulation or dispersion in which phase separation or agglomeration does not occur. (6) The objective of the composition of the invention claimed herein is to provide a suitable dielectric removal rate while preventing undesirable surface defects and ensuring high surface quality.

For a person having ordinary knowledge in the technical field to which the invention belongs, other objectives, advantages and applications of the invention claimed in this case will become obvious based on the following implementation methods.

The following embodiments are merely illustrative in nature and are not intended to limit the invention or the application and use of the invention claimed in this case. In addition, there is no intention to be limited by any theory appearing in the previous technical field, prior art, invention content or the following embodiments.

As used herein, the terms "comprising", "comprises" and "comprised of" are synonymous with "including", "includes" or "containing", "contains" and are inclusive or open-ended and do not exclude additional unlisted members, elements or method steps. It should be understood that as used herein, the term "comprising" includes the terms "consisting of", "consists" and "consists of".

In addition, in the specification and application, the terms "(A)", "(B)", "(C)", "(D)", etc. and the like are used to distinguish between similar elements and are not necessarily used to describe an order in a particular sequence or in a sequential order. It should be understood that such terms used in this manner are interchangeable under appropriate circumstances and that the embodiments of the claimed invention described herein are capable of being implemented in an order other than that described or illustrated herein. Where the terms “(A)”, “(B)” and “(C)” or “(a)”, “(b)”, “(c)”, “(d)”, “(i)”, “(ii)” etc. refer to steps of a method or use or analysis, there is no temporal or temporal coherence between the steps, i.e., the steps may be performed simultaneously or there may be seconds, minutes, hours, days, weeks, months or even years between the steps, unless otherwise indicated in this application as explained above or below herein.

In the following paragraphs, different aspects of the claimed invention are defined in more detail. Each aspect so defined may be combined with any other aspect or aspects unless expressly indicated to the contrary. In particular, any feature indicated as being preferred or advantageous may be combined with any other feature or features indicated as being preferred or advantageous.

Throughout this specification, reference to "one embodiment" or "an embodiment" or "a preferred embodiment" means that the particular features, structures or characteristics described in connection with the embodiment are included in at least one embodiment of the claimed invention. Therefore, the phrases "in one embodiment" or "in an embodiment" or "in a preferred embodiment" appearing throughout this specification do not necessarily all refer to the same embodiment but may refer to it. In addition, as would be apparent to a person of ordinary skill in the art to which the invention pertains based on this disclosure, features, structures or characteristics may be combined in any suitable manner in one or more embodiments. Furthermore, as will be appreciated by one of ordinary skill in the art to which the invention pertains, although some embodiments described herein include some but not other features included in other embodiments, combinations of features from different embodiments are intended to fall within the scope of the subject matter of the application and to be derived from different embodiments. For example, in the attached claims, any of the claimed embodiments may be used in any combination.

In addition, ranges defined throughout this specification also include extreme values, that is, a range of 1 to 10 means that 1 and 10 are both included in the range. For the avoidance of doubt, the applicant should be entitled to claim any equivalent under applicable law.

For the purposes of the claimed invention, "weight %" or "wt.%" as used in the claimed invention is relative to the total weight of the composition. Further, as described herein below, the sum of the wt.% of all compounds in the respective components totals 100 wt.-%.

For purposes of the claimed invention, a substrate is defined as a semiconductor wafer made of silicon or similar semi-metals used to manufacture microelectronic devices.

For the purposes of the claimed invention, polishing refers to the chemical mechanical removal of a specific layer from a substrate during a CMP process. The mechanical action is usually performed by a polishing pad, which is typically pressed against the surface to be polished and mounted on a moving platform. In a typical CMP process step, a wafer holder is rotated to bring the wafer to be polished into contact with the polishing pad. The CMP composition is usually applied between the wafer to be polished and the polishing pad.

For the purposes of the claimed invention, corrosion inhibitors are defined as chemical compounds that form a protective molecular layer on the surface of metals.

For the purposes of the claimed invention, a stabilizer is defined as a chemical compound that forms soluble complexes with iron (III) ions, deactivating the ions so that they cannot react normally with other elements or ions (such as silicates or phosphates) to produce precipitates or scale. It has been found that in the absence of a stabilizer, undesirable interactions between iron (III) oxidants and silicon-based compositions (such as the claimed invention) result in colloidal instability of the composition.

For the purposes of the claimed invention, an oxidizing agent is defined as a chemical compound that is capable of oxidizing the substrate to be polished or one of its layers.

For the purposes of the presently claimed invention, a pH adjuster is defined as a compound that is added to a composition to adjust its pH to a desired value.

For the purposes of the claimed invention, a buffer is defined as a component that resists pH changes following the addition of an acidic or alkaline component.

For the purposes of the claimed invention, the isoelectric point or pI is the pH at which an amino acid carries no net charge (Properties of Analytes and Matrices Determining HPLC Selection , Serban C. Moldoveanu , Victor David, in Selection of the HPLC Method in Chemical Analysis , 2017 ).

A lower dielectric constant increases the frequency at which a circuit can operate. Dielectric materials with lower dielectric constants used in metallization in IC manufacturing (primarily BEOL) are called low-k materials (dielectric constant k is, for example, 3.5 or lower) or ultra-low-k materials (dielectric constant k is, for example, 2.5 and lower). Several low-k materials are available from Applied Materials under the trade name Black Diamond (see generally US 6,974,777 B2 or Hosali et al., Analyzing damage from ultralow-k CMP, Solid State Technology, 48 (11) pg. 33, 2005). A detailed description of methods of making low-k materials and their deposition can be found in: McClatchie et al., Low Dielectric Constant Oxide Films Deposited Using CVD Techniques, DUMIC Conference Proceedings, (1998), pp. 311 et seq. For the purposes of the claimed invention, a low-k material is a material having a k value (dielectric constant) less than 3.5, preferably less than 3.0, and more preferably less than 2.7. An ultra-low-k material is a material having a k value (dielectric constant) less than 2.4.

For the purposes of the claimed invention, "colloidal silica" means silica that has been prepared by condensation polymerization of Si(OH) ₄ . The precursor Si(OH) ₄ can be obtained, for example, by hydrolysis of a high purity alkoxysilane or by acidification of an aqueous silicate solution. Such colloidal silica can be prepared according to US Pat. No. 5,230,833 or can be obtained in the form of any of a variety of commercially available products, such as Fuso® PL-1, PL-2 and PL-3 products and Nalco 1050, 2327 and 2329 products and other similar products available from DuPont, Bayer, Applied Research, Nissan Chemical, Nyacol and Clariant.

For the purposes of the claimed invention, "particle size" and "average particle size" are used interchangeably. The average particle size is defined as the _d50 value of the particle size distribution of the colloidal silica particles (A) in the aqueous medium (F).

For the purposes of the claimed invention, the term "substantially free" means that the composition does not contain any concentration of the component that may affect the polishing functionality of the composition. Preferably, the component is less than 1 ppm, more preferably less than 0.1 ppm, and most preferably less than the detection limit. For example, preferably the composition is substantially free of polyacrylamide and/or alkali metals such as potassium, and preferably the concentration of polyacrylamide and/or alkali metals such as potassium in the composition is less than 1 ppm. However, any trace amounts of such components that may remain on the semiconductor surface in the form of part of the residues from the pre-CMP process step will not be detrimental to the polishing application involving the composition described herein.

For the purposes of the claimed invention, the average particle size is measured, for example, using dynamic light scattering (DLS) or static light scattering (SLS) methods. These and other methods are well known in the art to which the invention belongs, see, for example, Kuntzsch, Timo; Witnik, Ulrike; Hollatz, Michael Stintz; Ripperger, Siegfried; Characterization of Slurries Used for Chemical-Mechanical Polishing (CMP) in the Semiconductor Industry; Chem. Eng. Technol; 26 (2003), Vol. 12, p. 1235. The results of such measurements are generally referred to in the literature as secondary particle size. Among these methods, DLS is the preferred method.

For the purposes of the claimed invention, for dynamic light scattering (DLS), a Malvern Zetasizer ZSP or a Horiba LB-550 V (DLS, dynamic light scattering measurement) or any other such instrument is typically used. This technique measures the hydrodynamic diameter of a particle when it scatters a laser light source (e.g. λ = 650 nm), detected at an angle of, for example, 90° or 173° relative to the incident light. Changes in the intensity of the scattered light are due to the random Brownian motion of the particles as they move through the incident light beam and are monitored as a function of time. The decay constant is extracted using the self-correlation function carried out by the instrument as a function of the delay time; smaller particles move through the incident beam at higher speeds and correspond to faster decay.

For the purposes of the claimed invention, the attenuation constant is proportional to the diffusion coefficient (D _t ) of the inorganic abrasive particles and is used to calculate the particle size according to the Stokes-Einstein equation: It is assumed that the suspended particles (1) have a spherical morphology and (2) are uniformly dispersed throughout the aqueous medium (i.e., do not agglomerate). This relationship is expected to be true for particle dispersions containing ≤ 1 wt% solids (solids concentration) since the viscosity of aqueous dispersions does not deviate significantly, with η = 0.96 mPa∙s (at T = 22 °C). The particle size distribution of aerosol or colloidal silica particle dispersions is typically measured in plastic cuvettes at 0.1 to 1.0 % solids concentration and dilutions, if necessary, in the dispersion medium or ultrapure water.

For the purposes of the claimed invention, the BET surface area of the colloidal silica particles is determined according to DIN ISO 9277:2010-09. The results of such a measurement are generally referred to in the literature as primary particle size.

For the purposes of the claimed invention, the disclosed measurement techniques are well known to those having ordinary knowledge in the art to which the invention belongs.

In the claimed invention, the dielectric polishing composition comprises: (A) surface-modified colloidal silica particles, the particles comprising negatively charged groups on the surface of the particles, wherein the surface-modified colloidal silica particles have a negative charge, a particle size from 60 nm to 200 nm, and a zeta potential of ≤ -35 mV at a pH ranging from ≥ 2.0 to ≤ 6.0; (B) at least one corrosion inhibitor selected from at least one guanidine derivative; (C) at least one iron (III) oxidizing agent; (D) at least one buffer selected from at least one basic amino acid having an isoelectric point (pI) ≥ 6.5; (E) at least one stabilizer; and (F)   an aqueous medium, wherein the pH of the composition ranges from ≥ 2.0 to ≤ 4.3, and wherein the polyacrylamide content of the composition is ≤ 1 ppm.

The composition has a polyacrylamide content of ≤ 1 ppm or in other words, the composition is substantially free of polyacrylamide. In some cases, a polyacrylamide content of > 1 ppm may have a negative impact on semiconductor topology. More preferably, the composition has a polyacrylamide content of ≤ 0.5 ppm, and even more preferably ≤ 0.1 ppm. For the purposes of the invention claimed in this case, the composition is substantially free of polyacrylamide and polyacrylamide copolymers. The copolymers of polyacrylamide can be cationic, anionic or non-ionic polyacrylamide copolymers. The presence of polyacrylamide copolymers in the composition results in undesirable silicon oxide (dielectric) depressions. Optimally, the amount of polyacrylamide or polyacrylamide copolymers in the composition is ≤ 0.01 ppm.

The composition is substantially free of alkali metals. Preferably, the composition contains ≤ 1 ppm of alkali metals. Alkali metals refer to elements of Group I of the Periodic Table, preferably selected from lithium, sodium, potassium, cadmium, cesium or cobalt, more preferably the composition is substantially free of lithium, sodium and potassium, even more preferably substantially free of sodium or potassium, and most preferably substantially free of potassium.

Alkaline metals such as sodium or potassium are typically introduced into the CMP composition through the use of acid/base or buffers, however, their presence can interfere with electronic devices on the wafer surface and have a detrimental effect on the final performance. Preferably, the amount of sodium or potassium in the composition is ≤ 1 ppm, even more preferably ≤ 0.1 ppm, and most preferably ≤ 0.01 ppm.

The composition is substantially free of phosphoric acid or its salts. The composition is substantially free of phosphates, such as organic phosphates or alkaline metal phosphates, such as potassium dihydrogen phosphate or sodium dihydrogen phosphate. Preferably, the composition contains ≤ 1 ppm of phosphoric acid or its salts. Phosphoric acid or its salts can combine with metals to form undesirable salts with low water solubility. Therefore, their presence can result in the need for an additional filtering step, thereby increasing the cost of the process.

The dielectric polishing composition of the present invention comprises components (A), (B), (C), (D), (E) and water and other components as required, as described below. (A) Surface modified colloidal silica particles

According to the claimed invention, the composition comprises negatively charged groups on the surface of the particles, wherein the surface modified colloidal silica particles have a negative charge, a particle size ranging from 60 nm to 200 nm, and a zeta potential of < -35 mV at a pH ranging from ≥ 2.0 to ≤ 6.0.

Typically, known silica selective compositions utilize niobium oxide, however, such abrasives have been noted to result in unacceptable matte finishes. Similar undesirable observations have been noted when utilizing fumed silica.

Preferably, the surface-modified silica particles have a zeta potential of < -35 mV, more preferably < -36 mV, even more preferably < -37 mV, and most preferably < -38 mV at a pH ranging from ≥ 2.0 to ≤ 6.0.

Preferably, the surface-modified silica particles have a zeta potential of > -50 mV, more preferably > -45 mV, and most preferably > -40 mV at a pH ranging from ≥ 2.0 to ≤ 6.0.

Preferably, the surface-modified silica particles have a zeta potential from -50 mV to -35 mV, more preferably from -40 mV to -35 mV, even more preferably from -45 mV to -35 mV, and most preferably from -39 mV to -35 mV at a pH ranging from ≥ 2.0 to ≤ 6.0.

The surface-modified silica particles are preferably amorphous and non-agglomerated and thus typically exist in the form of separate spheres that are not cross-linked with each other and contain hydroxyl groups on the surface. The surface-modified colloidal silica particles can be obtained by methods known in the art to which the invention belongs (such as ion exchange of silicates) or by sol-gel techniques (for example, hydrolysis or condensation of metal alkoxides or gelation of precipitated hydrated silica, etc.).

It is known that silica particles are stabilized by a permanent charge on their surface to prevent agglomeration and to ensure colloidal stability. The charge can be positive or negative. A negative charge is considered necessary because defects are observed on the surface of a substrate when positively charged (or cationic) silica particles are used for polishing (see FIG. 2 b). This is depicted in FIG. 2 a for a defect-free surface where a substrate is planarized with a composition comprising negatively charged (or anionic) silica particles. The charge on the surface is expressed as the zeta potential. The zeta potential of silica particles (without surface functionalization) depends on the pH of the aqueous medium (see FIG. 1 ). At a pH above 8, the zeta potential is equal to or below – 35 mV, i.e. low enough to ensure colloidal stability. Without being bound by theory, it is believed that at acidic pH (e.g., pH 2-6), the charge on the surface decreases (zeta potential is typically between +10 and -10 mV, for example) due to the interaction of silica with protons of the medium. Typical correlations of zeta potential and pH can be found in the literature (Esumi et al., Bull. Chem. Soc. Jpn., Vol. 61, 1988). The low surface charge may lead to the formation of aggregates as soon as shear forces are increased (e.g., by filtering or polishing).

The zeta potential may be affected by the presence of additives. However, as will be readily appreciated by a person of ordinary skill in the art to which the invention pertains, one or more of the mentioned components having a negligible or no effect on the surface charge of the particles may have no significant effect on the zeta potential of the particles (A) and thus a zeta potential of < -35 mV may still be achievable despite their presence during the measurement. Preferably, the zeta potential measurements mentioned herein for component (A) are made in the absence of the other components, i.e. the surface modified silica particles (A) have a zeta potential of < -35 mV when measured alone or in the substantial absence of the other components (B) to (E). The measurements outlined in FIGS. 1-4 have been made in the absence of the other components.

Preferably, the surface-modified colloidal silica particles (A) having a negative zeta potential of < -35 mV at a pH ranging from ≥ 2.0 to ≤ 6.0 are silica particles anionically modified with metallate ions or modified with sulfonic acid moieties.

As used herein, the term "anionic modified with metallate ions" refers in particular to silica particles in which metallate ions (i.e. M(OH) ₄ ^- ) are incorporated into the surface of the silica particles to replace Si(OH) ₄ sites and generate a permanent negative charge, as explained in WO 2006/028759 A2.

Preferably, the surface-modified colloidal silica particles (A) having a negative zeta potential of < -35 mV at a pH ranging from ≥ 2.0 to ≤ 6.0 are silica particles anionically modified with metallate ions. More preferably, the metallate ions are selected from aluminate, stannate, zincate or leadate.

Even more preferably, the surface-modified colloidal silica particles (A) having a negative zeta potential of < -35 mV at a pH ranging from ≥ 2.0 to ≤ 6.0 are silica particles modified with aluminate anions. Such surface-modified colloidal silica particles are disclosed in, for example, WO 2006/7028759 A2.

More preferably, the surface-modified colloidal silica particles of the component (A) having a negative zeta potential of < -35 mV at a pH ranging from ≥ 2.0 to ≤ 6.0 are silica particles anionically modified with sulfonic acid. Aqueous anionic silica sols modified with sulfonic acid that are highly stable under acidic conditions are disclosed, for example, in WO 2010734542 A1. Here, the aqueous anionic silica sol modified with sulfonic acid is obtained by a method in which a silane coupling agent having a functional group that can be chemically converted into a sulfonic acid group is chemically adsorbed onto the colloidal silica and then the functional group is converted into a sulfonic acid group. A preferred type of silica dispersion to be used for this type of chemical reaction is a silica dispersion with zeta potential functionality as shown in FIG. 1 (unmodified silica particles), where the charge on the surface of the particles is low in an acidic system. The same is considered a sign that the surface of the silica is clean and that the silane can easily react with the groups on the surface of the silica. If the charge on the silica particles is already high (e.g., -36 or -50 mV) before the silane coupling agent is charged to the reaction in an acidic system (e.g., pH 2-3), it is understood to be a sign that the surface of the silica is not clean and has been modified. The silane coupling agent may not be able to fully cover the surface and the charge on the surface may then remain low after the reaction and transformation. Surface modified particle dispersions with a zeta potential of, for example, pH 2-6, e.g. -23 mV, may have a low colloidal stability. This means that the dispersion may be easily destabilized by processes such as filtration or CMP in which shear forces occur. In such cases, filtered silica dispersions measured by DLS methods, e.g. with a Malvern Zetasizer ZSP (see FIG. 3 ), provide a signal with high variation, which cannot be easily interpreted as an average particle size. Other methods, such as deposition methods, have to be used here to assess the colloidal stability. On the other hand, it is noted that silica particles modified with sulfonic acid anions produce a stable DLS signal that can be easily measured (see FIG. 4 ).

Preferably, the concentration of the surface-modified colloidal silica particles (A) ranges from ≥ 3.2 wt.% to ≤ 13.0 wt.%, based on the total weight of the composition. The concentration of the surface-modified colloidal silica particles (A) is preferably not more than 13.0 wt.%, more preferably not more than 10.0 wt.%, particularly not more than 9.5 wt.%, even more preferably not more than 9.0 wt.%, even more preferably not more than 8.5 wt.%, even more preferably not more than 8.0 wt.%, for example not more than 7.5 wt.%, based on the total weight of the composition. It has been observed that a particle concentration exceeding 10 wt.% results in colloid instability in the composition. The concentration of the surface-modified colloidal silica particles (A) is preferably at least 3.2 wt.%, more preferably at least 3.5 wt.%, even more preferably at least 3.8 wt.%, particularly at least 4.0 wt.%, even more preferably at least 4.1 wt.%, still more preferably at least 4.2 wt.%, more preferably at least 4.3 wt.%, even more preferably at least 4.5 wt.%, still more preferably at least 4.8 wt.%, and most preferably at least 5.2 wt.%, based on the total weight of the composition. It was observed that a particle concentration below 3.2 wt.% resulted in undesirably high tungsten dishing and also resulted in low dielectric removal selectivity, i.e., a low silicon oxide to tungsten removal ratio. The concentration of the surface modified colloidal silica particles (A) is preferably in the range of ≥ 4.5 wt.% to ≤ 8.5 wt.% based on the total weight of the composition.

The surface-modified colloidal silica particles (A) are preferably contained in the composition in various particle size distributions. The particle size distribution of the surface-modified colloidal silica particles (A) can be unimodal or multimodal. In the case of multimodal particle size distributions, bimodal particle size distributions are often preferred. For the purposes of the claimed invention, unimodal particle size distributions are preferred for the surface-modified colloidal silica particles (A).

According to the claimed invention, the average particle diameter of the surface-modified colloidal silica particles (A) ranges from 60 nm to 200 nm as measured by dynamic light scattering technology. The mean or average particle diameter of the surface-modified colloidal silica particles (A) can vary within a wide range. The average particle size of the surface modified colloidal silica particles (A) is preferably in the range of ≥ 60 nm to ≤ 190 nm, more preferably in the range of ≥ 60 nm to ≤ 180 nm, more preferably in the range of ≥ 62 nm to ≤ 150 nm, more preferably in the range of ≥ 65 nm to ≤ 140 nm, particularly preferably in the range of ≥ 68 nm to ≤ 130 nm, and most preferably in the range of ≥ 70 nm to ≤ 120 nm, in each case measured by dynamic light scattering technique using an instrument such as a Zetasizer ZSP or a High Performance Particle Sizer (HPPS) from Malvern Instruments, Ltd. or a Horiba LB550.

The surface-modified colloidal silica particles (A) preferably have a variety of shapes. Thus, the particles (A) preferably have one or essentially only one type of shape. However, it is also possible that the particles (A) have different shapes. For example, there may be two types of particles (A) of different shapes. For example, (A) may have the following shapes: agglomerates, cubes, cubes with beveled edges, octahedrons, icosahedrons, coils, nodules or spheres, with or without protrusions or indentations.

Preferably, the surface-modified colloidal silica particles (A) are spherical, coiled, or a mixture of spherical and coiled particles. The spherical particles may or may not have protrusions or indentations. The coiled particles may or may not have protrusions or indentations. The coiled particles are preferably particles having a short axis of ≥ 10 nm to ≤ 200 nm and a major axis/minor axis ratio of preferably ≥ 1.4 to ≤ 2.2, more preferably ≥ 1.6 to ≤ 2.0. Preferably, the average shape factor is from ≥ 0.7 to ≤ 0.97, more preferably from ≥ 0.77 to ≤ 0.92, the average sphericity is from ≥ 0.4 to ≤ 0.9, more preferably from ≥ 0.5 to ≤ 0.7, and the average equivalent circle diameter is from ≥ 41 nm to ≤ 66 nm, more preferably from ≥ 48 nm to ≤ 60 nm, in each case determined by transmission electron microscopy and scanning electron microscopy.

Preferably, the surface-modified colloidal silica particles (A) are spherical or substantially spherical. The major axis/minor axis ratio of the spherical or substantially spherical particles is ≥ 0.9.

For the purpose of the claimed invention, the following explains the determination of the shape factor, sphericity and equivalent circular diameter of coiled particles. The shape factor provides information about the shape and indentation of individual particles and can be calculated according to the following formula: Shape factor = 4π (area / perimeter2)

The shape factor of a spherical particle with no indentation is 1. The value of the shape factor decreases as the amount of indentation increases. Sphericity provides information about the elongation of individual particles using moments about the mean and can be calculated as follows, where M is the central moment of gravity of the individual particle: Sphericity = (Mxx – Myy)-[4 Mxy2 + (Myy-Mxx)2]0.5 / (Mxx – Myy)+[4 Mxy2 + (Myy-Mxx)2]0.5 Elongation = (1 / Sphericity)0.5 Where Mxx = Σ (x-x mean)² /N Myy = Σ (y-y mean)² /N Mxy = Σ [(x-x mean)*(y-y mean)] /N N    Number of pixels forming the image of the individual particle x, y      Coordinates of the pixels x mean    Mean of the x coordinates of the N pixels forming the image of the particle y mean   The mean of the y coordinates of the N pixels that form the image of the particle.

The sphericity of spherical particles is 1. The value of the sphericity decreases when the particle is elongated. The equivalent circular diameter (hereinafter also abbreviated as ECD) of the individual non-spherical particles provides information about the diameter of a circle with the same area as the individual non-spherical particles. The average shape factor, average sphericity and average ECD are the arithmetic means of the individual properties related to the number of particles analyzed.

For the purposes of the invention claimed herein, the procedure used to define the particle shape characteristics is as follows. An aqueous dispersion of spherical silica particles having a solid content of 20 wt.% is dispersed on a carbon foil and dried. The dried dispersion is analyzed by using energy filtered transmission electron microscopy (EF-TEM) (120 kV) and scanning electron microscopy secondary electron imaging (SEM-SE) (5 kV). EF-TEM images with a resolution of 2k, 16 bits, and 0.6851 nm/pixel are used for analysis. The images are thresholded and binary encoded after noise suppression. Thereafter, the particles are separated manually. Overlapping and edge particles are identified and not used for analysis. ECD, shape factor, and sphericity as defined above are calculated and statistically classified. (B) Corrosion Inhibitors

According to the claimed invention, the composition comprises at least one corrosion inhibitor selected from at least one guanidine derivative.

As can be observed from Table 1 below, the corrosion inhibitor selected from at least one guanidine derivative prevents undesirable tungsten corrosion and high tungsten pitting.

Preferably, the at least one guanidine derivative is selected from buformin, phenformin, guanidine, proguanil hydrochloride, 2-guanidine benzimidazole, polyhexamethylene biguanide hydrochloride, polyaminopropyl biguanide, lohexidine or lohexidine salt.

More preferably, the guanidine derivative is selected from lohexidine or a lohexidine salt. It is known that lohexidine or its salt degrades into chemical sub-species over time, however, it is noted that the presence of the degradation itself or degradation products in the composition has little or no effect on the CMP activity outlined herein. The concentration of the degradation products will vary with the conditions (temperature/pressure, etc.). Some of the possible degradation products outlined in Table II-1 on page 19 of the paper entitled Studies on the mechanisms of solid state and solution instability of drugs by Zhixin Zong are reproduced below. Manipulation of the concentration of the degradation products and/or the like to enhance activity would be considered routine for one of ordinary skill in the art to which the invention pertains. MW Proposed structure

Another suitable guanidine derivative is alexidine ( N ¹ , N ^1′ -(hexane-1,6-diyl)bis[ N ³ -(2-ethylhexyl)iminodimethylamide])

Preferably, the lohexidine or lohexidine salt comprises the degradation products listed herein above.

Preferably, the corrosion inhibitor (B) is lohexidine.

Preferably, the corrosion inhibitor (B) is selected from lohexidine salts.

More preferably, the lohexidine salt is selected from the group consisting of lohexidine gluconate, lohexidine digluconate, lohexidine hydrochloride, lohexidine dihydrochloride, lohexidine acetate, lohexidine diacetate, lohexidine hexametaphosphate, lohexidine metaphosphate and lohexidine trimetaphosphate.

Most preferably, the corrosion inhibitor is selected from the group consisting of lohexidine, lohexidine gluconate and lohexidine digluconate.

Preferably, the corrosion inhibitor (B) is present in an amount ranging from ≥ 0.001 wt.% to ≤ 0.05 wt.%, based on the total weight of the composition. More preferably, the corrosion inhibitor (B) is present in an amount not exceeding 0.04 wt.%, and most preferably not exceeding 0.03 wt.%, based on the total weight of the composition. The amount of (B) is preferably at least 0.002 wt.%, and more preferably at least 0.003 wt.%, based on the total weight of the composition. The concentration of the corrosion inhibitor (B) is more preferably in the range of ≥ 0.003 wt.% to ≤ 0.03 wt.%, based on the total weight of the composition. (C) Iron (III) Oxidant

According to the claimed invention, the composition comprises at least one iron (III) oxidizing agent (C).

As can be observed from Table 1 below, the iron(III) oxidizing agent (C) oxidizes the substrate to be polished or one of its layers, thus ensuring a chemical contribution to the removal rate and high surface quality.

Preferably, the iron (III) oxidizing agent (C) is selected from iron (III) salts or compounds with nitric acid, sulfuric acid, hydrofluoric acid, hydrochloric acid, hydrobromic acid, hydroiodic acid, acetic acid, o-phosphoethanolamine, isophosphorous acid or mixtures thereof.

More preferably, the iron(III) oxidizing agent (C) is selected from iron(III) nitrate or its hydrate. Even more preferably, the iron(III) oxidizing agent (C) is iron(III) nitrate.

Preferably, the concentration of the iron (III) oxidant (C) ranges from ≥ 0.003 wt.% to ≤ 0.1 wt.%, based on the total weight of the composition. More preferably, the iron (III) oxidant (C) is present in an amount not exceeding 0.08 wt.%, even more preferably not exceeding 0.07 wt.%, optimally not exceeding 0.05 wt.%, optimally not exceeding 0.03 wt.%, based on the total weight of the composition. The amount of (C) is preferably at least 0.0035 wt.%, more preferably at least 0.004 wt.%, optimally at least 0.0045 wt.%, based on the total weight of the composition. When the amount of (C) is below < 0.003 wt%, undesirably high dielectric depression is noted. On the other hand, when the amount of (C) is > 0.1 wt.%, undesirably high tungsten MRR and depression are observed. The concentration of the iron (III) oxidizing agent (C) is preferably in the range of ≥ 0.0045 wt.% to ≤ 0.03 wt.%, and optimally in the range of ≥ 0.0048 wt.% to ≤ 0.02 wt.%, based on the total weight of the composition. (D) Buffer

According to the claimed invention, the composition comprises at least one buffer selected from at least one basic amino acid having an isoelectric point (pI) ≥ 6.5.

As can be observed from Table 1 below, the buffer (D) specifically prevents undesired recessing of the dielectric layer during chemical mechanical polishing while allowing for suitable maintenance of pH and high silicon oxide (dielectric) removal rates.

Chemically, buffers consist of a weak acid and its conjugated base or a weak base and its conjugated acid. For example, lysine, which has an _-NH2 group (a strong base), and its protonated ammonium form (having an _-NH3 ⁺ group or a conjugated acid form) combine to provide a buffer when in solution (such as when present in an aqueous composition of the claimed invention at a pH range of ≥ 2.0 to ≤ 6.0). It is noted that amino acids have a buffering range of ±1 pH unit from their pKa values. For the claimed invention, this would be around the pKa1 of the corresponding amino acid (lysine, arginine and histidine pKa1 ~ 2.18, 2.17 and 1.82 respectively, see: Carey and Giuliano (2011) Amino acids, peptides and proteins . Organic Chemistry 8th ed., 25, 1126 McGraw Hill, ISBN-13: 978-0077354770).

For the basic amino acids mentioned herein, their isoelectric points (pI) can be conveniently calculated by averaging the pKa values of the two amine groups (the two least acidic pKa values). For example, in the case of lysine, the pKa values of the amine groups are 8.95 and 10.53 and the calculated pI is 9.74. Similarly, the pI of arginine is 10.76 and the pI of histidine is 7.59. In addition, the pKa value can be obtained by titrating the group with a suitable acid/base. Preferably, the buffer (D) is a basic amino acid with an isoelectric point (pI) ≥ 6.9, more preferably ≥ 7.0, even more preferably ≥ 7.2, and most preferably ≥ 7.5.

Preferably, the buffer (D) is a basic amino acid selected from lysine, arginine or histidine. More preferably, the buffer (D) is a basic amino acid selected from arginine or histidine, and most preferably, the buffer (D) is arginine.

It is known that amino acids exist in the form of L or R optical isomers, of which the L form is biologically relevant and common. For the purpose of the invention claimed in this case, both isomers perform similar functions and can be used, however, for economic reasons, the L form may be preferred. Preferably, the basic amino acid is selected from L or R isomers.

Preferably, the concentration of the buffer (D) ranges from ≥ 0.1 wt.% to ≤ 0.78 wt.%, based on the total weight of the composition. More preferably, the buffer (D) is present in an amount not exceeding 0.75 wt.%, even more preferably not exceeding 0.73 wt.%, and most preferably not exceeding 0.7 wt.%, based on the total weight of the composition. The presence of (D) in an amount > 0.78 wt.% results in an undesirable increase in silicon oxide (dielectric) recessing. On the other hand, the presence of (D) in an amount < 0.1 wt.% results in insufficient pH stabilization. The amount of (D) is preferably at least 0.15 wt.%, more preferably at least 0.2 wt.%, even more preferably at least 0.25 wt.%, and most preferably at least 0.28 wt.%, based on the total weight of the composition. The concentration of the buffer (D) is more preferably in the range of ≥ 0.15 wt.% to ≤ 0.75 wt.%, and most preferably in the range of ≥ 0.25 wt.% to ≤ 0.73 wt.%, based on the total weight of the composition.

In a preferred embodiment, the buffer (D) is selected from histidine or arginine and is present in an amount of ≥ 0.15 wt.% to ≤ 0.75 wt.% based on the total weight of the composition.

In another preferred embodiment, the buffer (D) is arginine and is present in an amount of ≥ 0.15 wt.% to ≤ 0.75 wt.% based on the total weight of the composition. (E) Stabilizer

According to the claimed invention, the composition comprises at least one stabilizer (E).

Without wishing to be bound by theory, it is expected that agglomerate-formation is highly prevalent when using incompatible components such as silica particles, iron (III) salts and others. However, as can be observed from Table 1 below, the stabilizer (E) ensures colloidal stability.

Preferably, the at least one stabilizer (E) is selected from acetic acid, acetylacetonate, o-phosphoethanolamine, isophosphite, alendronic acid, acetic acid, phthalic acid, citric acid, adipic acid, oxalic acid, malonic acid, succinic acid, glutaric acid, pimelic acid, suppository acid, azelaic acid, sebacic acid, oxalic acid, citric acid, gluconic acid, muconic acid, ethylenediaminetetraacetic acid, propylenediaminetetraacetic acid, N,N-bis(carboxymethyl)alanine, nitrilotriacetic acid, diethylenetriaminepentaacetic acid, di(salicylidene)ethylenediamine, aminotris(methyleneisophosphite), diethylenetriaminepenta(methylisophosphite), ethylenediaminetetra(methyleneisophosphite) or a mixture thereof.

More preferably, the at least one stabilizer (E) is selected from phthalic acid, citric acid, adipic acid, oxalic acid, succinic acid, glutaric acid, pimelic acid, supemetic acid, azelaic acid, sebacic acid, oxalic acid, citric acid, gluconic acid, muconic acid, ethylenediaminetetraacetic acid, propylenediaminetetraacetic acid, N,N-bis(carboxymethyl)alanine, nitrilotriacetic acid, diethylenetriaminepentaacetic acid, di(salicylidene)ethylenediamine, aminotris(methyleneisophosphite), diethylenetriaminepenta(methylisophosphite), ethylenediaminetetra(methyleneisophosphite) or a mixture thereof.

Even more preferably, the at least one stabilizer (E) is selected from ethylenediaminetetraacetic acid, propylenediaminetetraacetic acid, N,N-di(carboxymethyl)alanine, nitrilotriacetic acid, diethylenetriaminepentaacetic acid, di(salicylidene)ethylenediamine, aminotris(methyleneisophosphite), diethylenetriaminepenta(methylisophosphite), ethylenediaminetetra(methyleneisophosphite) or a mixture thereof.

Most preferably, the at least one stabilizer (E) is ethylenediaminetetraacetic acid.

Preferably, the concentration of the stabilizer (E) ranges from ≥ 0.005 wt.% to ≤ 0.15 wt.% based on the total weight of the composition. More preferably, the stabilizer (E) is present in an amount not exceeding 0.1 wt.%, even more preferably not exceeding 0.08 wt.%, and most preferably not exceeding 0.03 wt.%, based on the total weight of the composition. The amount of (E) is preferably at least 0.0055 wt.%, more preferably at least 0.006 wt.%, and most preferably at least 0.008 wt.%, based on the total weight of the composition. The concentration of the stabilizer (E) is preferably in the range of ≥ 0.0055 wt.% to ≤ 0.08 wt.%, and the most preferably in the range of ≥ 0.008 wt.% to ≤ 0.03 wt.%, based on the total weight of the composition. (F) Aqueous Medium

According to the claimed invention, the composition comprises an aqueous medium (F). The aqueous medium (F) can be one type or a mixture of different types of aqueous media.

The aqueous medium (F) is preferably any medium containing water. Preferably, the aqueous medium (F) is a mixture of water and an organic solvent miscible with water. Representative examples of organic solvents include (but are not limited to) _C1 to _C3 alcohols, alkanediols and alkanediol derivatives.

More preferably, the aqueous medium (F) is water. In a preferred embodiment of the invention, the aqueous medium (F) is deionized water.

For the purposes of the claimed invention, if the amount of components other than (F) is y wt.% t of the composition, then the amount of (F) is (100 – y) wt.% of the composition.

The amount of the aqueous medium (F) in the composition is preferably not more than 99.9 wt.%, more preferably not more than 99.6 wt.%, most preferably not more than 99 wt.%, particularly preferably not more than 98 wt.%, especially not more than 97 wt.%, for example not more than 95 wt.%, based on the total weight of the composition. The amount of the aqueous medium (F) in the composition is preferably at least 65 wt.%, more preferably at least 75 wt.%, most preferably at least 85 wt.%, particularly preferably at least 88 wt.%, especially at least 90 wt.%, for example at least 92.5 wt.%, based on the total weight of the composition.

The properties of the composition can be based on the pH of the corresponding composition. According to the claimed invention, the pH of the composition ranges from ≥ 2.0 to ≤ 4.3. Preferably, the pH of the composition is ≤ 4.2, more preferably ≤ 4.1, optimally ≤ 4.05, particularly preferably ≤ 4.0, and most preferably ≤ 3.5. The pH of the composition is preferably ≥ 2.1, more preferably ≥ 2.3, optimally ≥ 2.5, particularly preferably ≥ 2.6, and most preferably ≥ 2.8. The pH of the composition preferably ranges from ≥ 2.1 to ≤ 4.2, preferably from ≥ 2.3 to ≤ 4.1, more preferably from ≥ 2.5 to ≤ 4.0, and most preferably from ≥ 2.8 to ≤ 3.5.

Preferably, the composition further comprises an additive selected from a pH adjuster, an oxidant, a wetting agent, a dispersant, a biocidal agent or a mixture thereof. More preferably, the at least one additive is different from components (A), (B), (C), (D), (E) and (F) and is added as needed in addition to the components.

Preferably, the at least one pH adjuster is selected from the group consisting of: inorganic acids, carboxylic acids, amine bases, ammonium hydroxides, including tetraalkylammonium hydroxides. Preferably, the at least one pH adjuster is selected from the group consisting of: nitric acid, sulfuric acid and ammonia. More preferably, the pH adjuster is nitric acid.

The amount of the at least one pH adjuster is preferably not more than 10 wt.%, more preferably not more than 2 wt.%, most preferably not more than 0.5 wt.%, especially not more than 0.1 wt.%, for example not more than 0.05 wt.%. The amount of the at least one pH adjuster is preferably at least 0.0005 wt.%, more preferably at least 0.005 wt.%, most preferably at least 0.025 wt.%, especially at least 0.1 wt.%, for example at least 0.4 wt.%, based on the total weight of the composition.

Preferably, the composition of the invention claimed in this case may further contain at least one oxidizing agent.

Preferably, the at least one oxidizing agent is selected from the group consisting of organic peroxides, inorganic peroxides, nitrates, persulfates, iodates, periodic acid, periodates, permanganates, perchloric acid, perchlorates, bromic acid and bromates. The oxidizing agent is optionally present in addition to the iron(III) oxidizing agent.

More preferably, the at least one oxidizing agent is hydrogen peroxide.

Preferably, the at least one oxidizing agent is present in an amount ranging from ≥ 0.01 wt.% to ≤ 1.0 wt.%, based on the total weight of the composition.

Preferably, the concentration of the at least one oxidizing agent is no more than 5.0 wt.%, even more preferably no more than 2.0 wt.%, even more preferably no more than 1.0 wt.%, even more preferably no more than 0.8 wt.%, and most preferably no more than 0.5 wt.%, in each case based on the total weight of the composition. The concentration of the at least one oxidizing agent is preferably at least 0.01 wt.%, more preferably at least 0.05 wt.%, and most preferably at least 0.1 wt.%, in each case based on the total weight of the composition.

More preferably, the concentration of hydrogen peroxide as the oxidizing agent is ≥ 0.01 wt.% to ≤ 1.0 wt.%, even more preferably ≥ 0.05 wt.% to ≤ 1.0 wt.%, most preferably ≥ 0.05 wt.% to ≤ 0.5 wt.%, particularly preferably ≥ 0.01 wt.% to ≤ 0.1 wt.%, in each case based on the total weight of the composition.

Methods for preparing compositions for chemical mechanical polishing are generally known. These methods can be applied to prepare the compositions of the claimed invention. This can be done by dispersing or dissolving the components (A), (B), (C), (D) and (E) described above in the aqueous medium (F), preferably water, and optionally by adjusting the pH by adding acids and/or bases (pH adjusters). For this purpose, customary and standard mixing methods and mixing equipment can be used, such as stirring tanks, high shear impellers, ultrasonic mixers, homogenizer nozzles or countercurrent mixers.

The preferred embodiment of the invention claimed in this case is directed to a composition comprising the following components: (A) surface-modified colloidal silica particles, the particles comprising negatively charged groups on the surface of the particles, wherein the surface-modified colloidal silica particles have a negative charge, a particle size from 60 nm to 200 nm, and a zeta potential of < -35 mV at a pH from ≥ 2.0 to ≤ 6.0; (B) at least one corrosion inhibitor, the corrosion inhibitor selected from at least one guanidine derivative; (C) at least one iron (III) oxidizing agent; (D) at least one buffer, the buffer selected from at least one basic amino acid having an isoelectric point (pI) ≥ 6.5; (E) At least one stabilizer selected from ethylenediaminetetraacetic acid, propylenediaminetetraacetic acid, N,N-di(carboxymethyl)alanine, nitrilotriacetic acid, diethylenetriaminepentaacetic acid, di(styrene)ethylenediamine, aminotris(methyleneisophosphite), diethylenetriaminepenta(methylisophosphite), ethylenediaminetetra(methyleneisophosphite) or a mixture thereof; and (F)   an aqueous medium, wherein the pH of the composition ranges from ≥ 2.0 to ≤ 4.3, and wherein the polyacrylamide content of the composition is ≤ 1 ppm.

Another preferred embodiment of the invention claimed in this case is directed to a composition comprising the following components: (A) Surface-modified colloidal silica particles, the particles comprising negatively charged groups on the surface of the particles, wherein the surface-modified colloidal silica particles have a negative charge, a particle size of from 60 nm to 200 nm, a zeta potential of < -35 mV at a pH of from ≥ 2.0 to ≤ 6.0, and a concentration ranging from ≥ 3.2 wt.% to ≤ 13.0 wt.% based on the total weight of the composition; (B)   At least one corrosion inhibitor, the corrosion inhibitor selected from at least one guanidine derivative and having a concentration ranging from ≥ 0.001 wt.% to ≤ 0.05 wt.% based on the total weight of the composition; (C)  At least one iron (III) oxidant, the iron (III) oxidant ranging from ≥ 0.003 wt.% to ≤ 0.1 wt.% based on the total weight of the composition; (D) At least one buffering agent, the buffering agent is selected from at least one basic amino acid with an isoelectric point (pI) ≥ 6.5 and the concentration ranges from ≥ 0.1 wt.% to ≤ 0.78 wt.% based on the total weight of the composition; (E) At least one stabilizer, the stabilizer is selected from ethylenediaminetetraacetic acid, propylenediaminetetraacetic acid, N,N-di(carboxymethyl)alanine, nitrilotriacetic acid, diethylenetriaminepentaacetic acid, di(styrene)ethylenediamine, aminotris(methyleneisophosphite), diethylenetriaminepenta(methylisophosphite), ethylenediaminetetra(methyleneisophosphite) or a mixture thereof and the concentration ranges from ≥ 0.005 wt.% to ≤ 0.15 wt.% based on the total weight of the composition; and (F)   an aqueous medium, wherein the pH range of the composition is from ≥ 2.0 to ≤ 4.3, and wherein the polyacrylamide content of the composition is ≤ 1 ppm.

Another preferred embodiment of the invention claimed in this case is directed to a composition comprising the following components: (A) surface-modified colloidal silica particles, the particles comprising negatively charged groups on the surface of the particles, wherein the surface-modified colloidal silica particles have a negative charge, a particle size of from 60 nm to 200 nm, a zeta potential of < -35 mV at a pH of from ≥ 2.0 to ≤ 6.0, and a concentration ranging from ≥ 3.2 wt.% to ≤ 13.0 wt.% based on the total weight of the composition; (B) at least one corrosion inhibitor, the corrosion inhibitor selected from at least one guanidine derivative, and a concentration ranging from ≥ 0.001 wt.% to ≤ 0.05 wt.% based on the total weight of the composition; (C)   At least one iron (III) oxidant, the iron (III) oxidant ranging from ≥ 0.003 wt.% to ≤ 0.1 wt.% based on the total weight of the composition; (D)  At least one buffering agent, the buffering agent selected from at least one basic amino acid having an isoelectric point (pI) ≥ 6.5, and the concentration ranging from ≥ 0.1 wt.% to ≤ 0.78 wt.% based on the total weight of the composition; (E)  At least one stabilizer, the stabilizer is selected from ethylenediaminetetraacetic acid, propylenediaminetetraacetic acid, N,N-di(carboxymethyl)alanine, nitrilotriacetic acid, diethylenetriaminepentaacetic acid, di(styrene)ethylenediamine, aminotris(methyleneisophosphite), diethylenetriaminepenta(methylisophosphite), ethylenediaminetetra(methyleneisophosphite) or a mixture thereof, and the concentration ranges from ≥ 0.005 wt.% to ≤ 0.15 wt.% based on the total weight of the composition; and (F)   an aqueous medium, wherein the pH range of the composition is from ≥ 2.0 to ≤ 4.3, wherein the polyacrylamide content of the composition is ≤ 1 ppm, and wherein the alkali metal content of the composition is ≤ 1 ppm.

Another preferred embodiment of the invention claimed in this case is directed to a composition comprising the following components: (A) Surface-modified colloidal silica particles, the particles comprising negatively charged groups on the surface of the particles, wherein the surface-modified colloidal silica particles have a negative charge, a particle size of from 60 nm to 200 nm, a zeta potential of < -35 mV at a pH of from ≥ 2.5 to ≤ 6.0, and a concentration ranging from ≥ 0.3 wt.% to ≤ 7.0 wt.% based on the total weight of the composition; (B)   At least one corrosion inhibitor, the corrosion inhibitor selected from at least one guanidine derivative, and the concentration ranging from ≥ 0.003 wt.% to ≤ 0.03 wt.% based on the total weight of the composition; (C)  At least one iron (III) oxidant, the iron (III) oxidant ranging from ≥ 0.0048 wt.% to ≤ 0.02 wt.% based on the total weight of the composition; (D) At least one buffering agent, the buffering agent selected from at least one basic amino acid having an isoelectric point (pI) ≥ 6.5, and the concentration ranging from ≥ 0.15 wt.% to ≤ 0.75 wt.% based on the total weight of the composition; (E) At least one stabilizer, the stabilizer is selected from ethylenediaminetetraacetic acid, propylenediaminetetraacetic acid, N,N-di(carboxymethyl)alanine, nitrilotriacetic acid, diethylenetriaminepentaacetic acid, di(styrene)ethylenediamine, aminotris(methyleneisophosphite), diethylenetriaminepenta(methylisophosphite), ethylenediaminetetra(methyleneisophosphite) or a mixture thereof, and the concentration ranges from ≥ 0.008 wt.% to ≤ 0.03 wt.% based on the total weight of the composition; and (F)   an aqueous medium, wherein the pH range of the composition is from ≥ 2.5 to ≤ 4.3, and wherein the polyacrylamide content of the composition is ≤ 1 ppm.

The preferred embodiment of the claimed invention is directed to a composition comprising: (A) surface-modified colloidal silica particles comprising negatively charged groups on the surface of the particles, wherein the surface-modified colloidal silica particles have a negative charge, a particle size of from 60 nm to 200 nm, a zeta potential of < -35 mV at a pH of from ≥ 2.5 to ≤ 6.0, and a concentration ranging from ≥ 0.3 wt.% to ≤ 7.0 wt.% based on the total weight of the composition; and (B) (C) at least one corrosion inhibitor, the corrosion inhibitor is selected from at least one guanidine derivative, the at least one guanidine derivative is selected from buformin, phenformin, guanidine, chlorguanidine hydrochloride, 2-guanidine benzimidazole, polyhexamethylene biguanide hydrochloride, polyaminopropyl biguanide, lohexidine or a lohexidine salt, preferably lohexidine or a lohexidine salt, and the concentration is in the range of ≥ 0.003 wt.% to ≤ 0.03 wt.% based on the total weight of the composition; (D) at least one iron (III) oxidant, the iron (III) oxidant is in the range of ≥ 0.0048 wt.% to ≤ 0.02 wt.% based on the total weight of the composition; at least one buffer (D) selected from lysine, arginine or histidine and having a concentration ranging from ≥ 0.15 wt.% to ≤ 0.75 wt.% based on the total weight of the composition; (E) at least one stabilizer selected from ethylenediaminetetraacetic acid, propylenediaminetetraacetic acid, N,N-di(carboxymethyl)alanine, nitrilotriacetic acid, diethylenetriaminepentaacetic acid, di(salicylidene)ethylenediamine, aminotris(methyleneisophosphite), diethylenetriaminepenta(methylisophosphite), ethylenediaminetetra(methyleneisophosphite) or a mixture thereof and having a concentration ranging from ≥ 0.008 wt.% to ≤ 0.03 wt.% based on the total weight of the composition; and (F) an aqueous medium, The pH of the composition ranges from ≥ 2.5 to ≤ 4.3, and the polyacrylamide content of the composition is ≤ 1 ppm. Process for manufacturing semiconductor devices

On the other hand, the claimed invention is directed to a process for manufacturing a semiconductor device comprising chemically mechanically polishing a substrate (S) used in the semiconductor industry in the presence of a composition as described herein, the substrate (S) comprising (i) tungsten and/or (ii) a tungsten alloy; and (iii) at least one dielectric layer selected from silicon, silicon oxide, silicon nitride or a low-k material.

Preferably, the dielectric layer is selected from silicon oxide, silicon nitride or a combination thereof.

Preferably, the ratio of silicon oxide material removal rate (MRR) to tungsten material removal rate (MRR) is < 15.0, more preferably < 14.5. Still more preferably, the ratio of silicon oxide material removal rate (MRR) to tungsten material removal rate (MRR) ranges from 3:1 to 15:1. Most preferably, the ratio of silicon oxide material removal rate (MRR) to tungsten material removal rate (MRR) ranges from 3.5:1 to 10:1. Without being bound by theory, high TEOS/silicon oxide to tungsten selectivity (low silicon oxide to tungsten ratio ≥ 15.0) may result in undesirable micro scratches or surface roughness.

Preferably, the SER is < 30 ppb. More preferably, the SER is < 25 ppb. Even better, the SER is < 23 ppb. Best, the SER is < 22 ppb.

Better, MRR > 300 Å/min for silicon oxide. Better, MRR > 350 Å/min for silicon oxide. Even better, MRR > 420 Å/min for silicon oxide. Best, MRR > 450 Å/min for silicon oxide.

Better, tungsten MRR < 200 Å/min. Better, tungsten MRR < 180 Å/min. Best, tungsten MRR < 130 Å/min.

The semiconductor components that can be manufactured by the process according to the invention claimed in this case are not particularly limited. The semiconductor components can be electronic components containing semiconductor materials (such as, for example, silicon, germanium and III-V materials). The semiconductor components can be manufactured in the form of a single separate component or in the form of an integrated circuit (IC) composed of several components manufactured and interconnected on a wafer. The semiconductor component can be a two-terminal component, such as a diode, a three-terminal component, such as a bipolar transistor, a four-terminal component, such as a Hall effect sensor, or a multi-terminal component. Preferably, the semiconductor component is a multi-terminal component. The multi-terminal component may be a logic component, such as an integrated circuit, and a microprocessor or a memory device, such as a random access memory (RAM), a read-only memory (ROM) and a phase-change random access memory (PCRAM). Preferably, the semiconductor component is a multi-terminal logic component. In particular, the semiconductor component is an integrated circuit or a microprocessor.

Typically, in integrated circuits, tungsten (W) is used for M0 or M1 interconnects. Excess tungsten on the dielectric can be removed by a known chemical mechanical polishing process.

Generally, the tungsten/tungsten alloy can be manufactured or obtained in different ways, such as ALD, PVD or CVD processes. Generally, the tungsten and/or tungsten alloy can be of any type, form or shape. Preferably, the tungsten and/or tungsten alloy has a layered and/or overgrowth shape. If the tungsten and/or tungsten alloy has a layered and/or overgrowth shape, the tungsten and/or tungsten alloy content is preferably more than 90% of the corresponding layer and/or overgrowth, more preferably more than 95%, most preferably more than 98%, especially more than 99%, for example more than 99.9% by weight. The tungsten and/or tungsten alloy is preferably filled or grown in trenches or pillars between other substrates, more preferably in dielectric materials such as, for example, _SiO2 , silicon, low-k (BD1, BD2) or ultra-low-k materials or other isolation and semiconductor materials used in the semiconductor industry. For example, in a through silicon via (TSV) intermediate process, insulating materials such as polymers, photoresists and/or polyimides can be used as insulating materials for insulating/isolation properties after the TSV is exposed from the back side of the wafer between subsequent wet etching and CMP processing steps. Uses

In another aspect, the claimed invention is directed to the use of the compositions described herein for polishing a substrate (S) comprising: (i) tungsten and/or (ii) a tungsten alloy; and (iii) at least one dielectric layer.

Preferably, the at least one dielectric layer is selected from silicon, silicon oxide, silicon nitride or a low-k material. More preferably, the dielectric layer comprises silicon oxide, silicon nitride or a combination thereof.

Preferably, the composition is used in semiconductor manufacturing and its process.

The compositions according to the claimed invention have at least one of the following advantages: (1) The compositions and methods of the claimed invention show high selectivity for the removal of silicon oxide over tungsten. (2) The compositions and methods of the claimed invention show improved performance in inhibiting etching, especially inhibiting the etching of tungsten and cobalt (as evidenced by low SER values). (3) The compositions of the claimed invention provide stable formulations or dispersions in which phase separation or aggregation does not occur, especially in acidic systems. (4) The compositions of the claimed invention allow for easy processing, such as compatibility with industrially relevant steps (such as microfiltration). (5) The processes of the claimed invention are easy to apply and require as few steps as possible. (6) The compositions and methods of the claimed invention allow for good tunability, thereby achieving high silicon oxide (SiO ₂ ) removal rates while ensuring low tungsten (W) removal rates. (7) The goal of the compositions of the claimed invention is to provide the above-mentioned suitable removal rates while preventing undesirable surface defects and ensuring high surface quality. ( 8 ) The compositions and methods of the claimed invention suppress undesirable recessing of the tungsten layer during chemical mechanical polishing while maintaining low levels of recessing of the dielectric layer.

Below, a list of implementations is provided to further illustrate the present disclosure but is not intended to limit the disclosure to the specific implementations listed below.

1. A dielectric polishing composition comprising: (A) surface modified colloidal silica particles comprising negatively charged groups on the surface of the particles, wherein the surface modified colloidal silica particles have a negative charge, a particle size of from 60 nm to 200 nm, and a zeta potential of ≤ -35 mV at a pH ranging from ≥ 2.0 to ≤ 6.0; (B) at least one corrosion inhibitor selected from at least one guanidine derivative; (C) at least one iron (III) oxidizing agent; (D) at least one buffer selected from at least one basic amino acid having an isoelectric point (pI) ≥ 6.5; (E) at least one stabilizer; and (F)   an aqueous medium, wherein the pH of the composition ranges from ≥ 2.0 to ≤ 4.3, and wherein the polyacrylamide content of the composition is ≤ 1 ppm.

2. A composition as in embodiment 1, wherein the surface modified colloidal silica particles have a zeta potential from -35 mV to -50 mV at a pH ranging from ≥ 2.0 to ≤ 6.0.

3.    A composition as in any of the preceding embodiments, wherein the concentration of the surface modified colloidal silica particles (A) ranges from ≥ 3.2 wt.% to ≤ 13.0 wt.% based on the total weight of the composition.

4. A composition as in any of the preceding embodiments, wherein the guanidine derivative is selected from buformin, phenformin, guanidine, chlorguanil hydrochloride, 2-guanidine benzimidazole, polyhexamethylene biguanide hydrochloride, polyaminopropyl biguanide, lohexidine or a lohexidine salt, preferably lohexidine or a lohexidine salt.

5. The composition of any of the preceding embodiments, wherein the concentration of the corrosion inhibitor (B) is in the range of ≥ 0.001 wt.% to ≤ 0.05 wt.% based on the total weight of the composition.

6. A composition as in any of the preceding embodiments, wherein the pH of the composition ranges from ≥ 3.0 to ≤ 4.5.

7. A composition as in any of the preceding embodiments, wherein the iron (III) oxidizing agent (C) is selected from iron (III) nitrate or a hydrate thereof.

8.    A composition as in any of the preceding embodiments, wherein the concentration of the iron(III) oxidant (C) is in the range of ≥ 0.003 wt.% to ≤ 0.1 wt.% based on the total weight of the composition.

9. A composition as in any of the preceding embodiments, wherein the buffer (D) is a basic amino acid selected from lysine, arginine or histidine.

10. A composition as in any of the preceding embodiments, wherein the concentration of the buffer (D) ranges from ≥ 0.1 wt.% to ≤ 0.78 wt.% based on the total weight of the composition.

11. A composition as in any of the preceding embodiments, wherein the stabilizer (E) is selected from acetic acid, acetylacetonate, o-phosphoethanolamine, isophosphite, alonphosphonic acid, acetic acid, phthalic acid, citric acid, adipic acid, oxalic acid, malonic acid, succinic acid, glutaric acid, pimelic acid, suppository acid, azelaic acid, sebacic acid, oxalic acid, citric acid, gluconic acid, muconic acid, ethylenediaminetetraacetic acid, propylenediaminetetraacetic acid, N,N-bis(carboxymethyl)alanine, nitrilotriacetic acid, diethylenetriaminepentaacetic acid, di(styrene)ethylenediamine, aminotris(methyleneisophosphite), diethylenetriaminepenta(methylisophosphite), ethylenediaminetetra(methyleneisophosphite) or a mixture thereof.

12. A composition as in any of the preceding embodiments, wherein the concentration of the stabilizer (E) ranges from ≥ 0.005 wt.% to ≤ 0.15 wt.% based on the total weight of the composition.

13. A composition as in any of the preceding embodiments, wherein the potassium content of the composition is ≤ 1 ppm.

14. A composition as in any of the preceding embodiments, wherein the composition further comprises an additive selected from a pH adjuster, an oxidant, a wetting agent, a dispersant, a biocide or a mixture thereof.

15. A composition as in any of the preceding embodiments, wherein the composition is used for polishing a substrate (S), wherein the substrate (S) comprises: (i) tungsten and/or (ii) a tungsten alloy; and (iii) at least one dielectric layer selected from silicon, silicon oxide, silicon nitride or a low-k material.

16. A process for manufacturing a semiconductor device, comprising chemically mechanically polishing a substrate (S) used in the semiconductor industry in the presence of a composition as defined in any of the preceding embodiments, wherein the substrate (S) comprises (i) tungsten and/or (ii) a tungsten alloy; and (iii) at least one dielectric layer selected from silicon, silicon oxide, silicon nitride or a low-k material.

17. A process as in embodiment 16, wherein the ratio of the material removal rate (MRR) of silicon oxide to the material removal rate (MRR) of tungsten ranges from 2:1 to 100:1.

18. A process as in any one of embodiments 16 to 17, wherein the static etch rate (SER) of tungsten is less than 30 ppb.

19. A process as in any one of embodiments 16 to 18, wherein the material removal rate (MRR) of silicon oxide is > 300 Å/min.

20. A process as in any of embodiments 16 to 19, wherein the material removal rate (MRR) of tungsten is less than 200 Å/min.

21. Use of a composition as described in any one of embodiments 1 to 15 for polishing a substrate (S) comprising: (i) tungsten and/or (ii) a tungsten alloy; and (iii) at least one dielectric layer.

22. The method of implementation 21, wherein the at least one dielectric is selected from silicon, silicon oxide, silicon nitride or low-k material.

Although the claimed invention has been described with respect to its particular embodiments, certain modifications and equivalents will be apparent to those skilled in the art and are intended to be included within the scope of the claimed invention .

The claimed invention is described in detail by the following embodiments. More specifically, the test methods explicitly described below are part of the general disclosure of the present application and are not limited to the specific embodiments.

The following describes the general procedure and experiment for preparing the slurry.

•     Silica particles, purchased from Fuso Chemical Corporation under the trade names Fuso® PLXC (cationic particles) and Fuso® PL5D (anionic particles)

•     Lohexidine and lohexidine digluconate, available from Sigma Aldrich

•     Deionized water, available from BASF SE

•     Polyacrylamide (Mn = 10000 g/mol), available from Sigma Aldrich

•     Ethylenediaminetetraacetic acid (EDTA), available from Sigma Aldrich

•     Iron nitrate nonahydrate, available from Sigma Aldrich

•     L-arginine, available from Sigma Aldrich

•     L-Histidine, available from Sigma Aldrich

•     L-Lysine, available from Sigma Aldrich

• L-Glycine, available from Sigma Aldrich as a slurry composition:

The slurry composition comprises: (A) silica particles with a particle size ranging from 60 nm to 200 nm and a zeta potential of < -35 mV at a pH ranging from ≥ 2.0 to ≤ 6.0 (B) corrosion inhibitor: lohexidine digluconate (C) oxidant: iron (III) nitrate (D) buffer: lysine, histidine or arginine; (E) stabilizer: ethylenediaminetetraacetic acid (EDTA); and (F) deionized water (DIW) Inorganic particles used in the method in the embodiment (A)

An embodiment according to the present invention comprises colloidal silica particles (A1) having an average secondary particle size (d2) of 109 nm (measured using a Malvern Zeta Sizer ZSP instrument using a dynamic light scattering (DLS) technique). The silica surface has been modified with sulfonic acid moieties. Procedure for defining particle shape characteristics

An aqueous dispersion of coiled silica particles with a solid content of 20 wt.% was dispersed on a carbon foil and dried. The dried dispersion was analyzed by using energy-filtered transmission electron microscopy (EF-TEM) (120 kV) and scanning electron microscopy secondary electron imaging (SEM-SE) (5 kV). EF-TEM images with a resolution of 2k, 16 bits, and 0.6851 nm/pixel were used for the analysis. The images were encoded using threshold binary encoding after noise suppression. Thereafter, the particles were separated manually. Overlapping and edge particles were identified and not used for analysis. ECD, shape factor, and sphericity as defined above were calculated and statistically classified. Measurement of zeta potential

Zeta potential values were measured with a Malvern Zetasizer ZSP (software version 7.11) equipped with a DTS1070 disposable folded capillary cell. The measurements were recorded at 25 °C and 0.1% solid concentration. To ensure this, the filtered sample solution (aqueous) was passed through a Millex SV low protein Durapore PVDF membrane (5 µm). The zeta potential was calculated from the measured electromobility law and the particle size (obtained from DLS measurements) and fitted to the Smoluchowski model. Particle size measurement - dynamic light scattering ( DLS )

The measurements were performed with a Malvern Zetasizer ZSP equipped with a semi-micro polystyrene cell. The measurements were performed at 25°C by dispersing the particles (A) in water (0.1%). Instrument settings: Dispersant: water (viscosity: 0.8872 mPa*s; RI 1.330); 5 measurements, 60 s each; Auto-attenuator selection: measurement position fixed at 4.65; Analysis mode: normal purpose. Filtered sample solution (aqueous) through a Millex SV low protein Durapore PVDF membrane (5 µm).

Figures 3 and 4 depict the DLS measurement results of the filtered silica dispersion measured by Malvern Zetasizer ZSP when the zeta potential of the surface-modified colloidal silica particles was > -35 mV and < -35 mV in the pH range from ≥ 2.0 to ≤ 6.0, respectively. Figure 3 shows that the DLS method cannot record stable measurement results when the zeta potential of the surface-modified colloidal silica particles is > -35 mV (zeta potential adjusted with KCl) in the pH range from ≥ 2.0 to ≤ 6.0. On the other hand, when the zeta potential of the surface-modified colloidal silica particles is < -35 mV in the pH range from ≥ 2.0 to ≤ 6.0, a stable signal can be recorded (see Figure 4). Procedure for preparing slurry composition

The components of the slurry composition were mixed thoroughly and all mixing procedures were carried out under stirring. Aqueous stock solutions of each of compounds (A), (B), (C), (D) and (E) were prepared by dissolving the desired amount of the respective compound in ultrapure water (UPW). The pH of the stock solutions was adjusted to the desired value using nitric/phosphoric acid. The stock solution of (B) had a concentration of 20 wt.% of Lohexidine digluconate solution and that of (C) 0.08 wt.%. For (A), a dispersion provided by the supplier was used, typically about 20% - 30% abrasive concentration by weight.

Alternatively, the oxidizing agent (C) can be used in the form of a Fe(III) EDTA solution obtainable by generally known methods, for example as reported by Lind et al., Stereochemistry of Ethylenediamintetraacetato Complexes , Inorganic Chemistry Vol . 3, No 1, 1964 (page 34 et seq.).

The final composition was passed through a 0.1 µm syringe filter prior to CMP. Filtration is an important process in the industry and the colloidal stability of the composition is further highlighted by its ability to remain colloidal despite microfiltration. pH Measurement

The pH value was measured using a pH electrode combination (Schott, blue line 22 pH electrode). Static Etching Rate ( SER ) Experiment

The SER experiment was conducted as follows: • The tungsten (W) coated wafer was cut into several 2.5x2.5 cm specimens and washed with deionized water (DIW). • Each specimen was treated with 0.1% citric acid solution for 4 min and then washed with DIW. • 300 ml of freshly prepared slurry was placed in a beaker and heated to 60 °C. • The tungsten (W) specimen was placed in the slurry and kept in the slurry for 10 min. in the SER equipment. • The tungsten (W) specimen was removed and rinsed with DIW for 1 min and dried with nitrogen. • The concentration of tungsten ions in the slurry was measured by ICP-MS after etching. Standard CMP process for barrier polished wafers:

The GnP POLI-500 polishing tool is used to planarize semiconductor films on sample-sized silicon wafers. This tool is available in several different configurations based on the user's needs for carrier type, pad, and conditioning. The CMP parameter settings on the GnP tool (such as polishing pressure, dressing pressure, slurry flow rate, carrier/stage rotation rate) are established before polishing to obtain the desired planarity and material removal rate.

Before the polishing process begins, condition the pad on the platform. Conditioning involves using a diamond disc to remove any debris or hardened material from the pad surface and restore its optimal texture. Place the sample wafer face down on a carrier with a special film holder that has a square groove in the center and is sized to match the sample wafer.

The CMP process begins with a rotating polishing pad in contact with the wafer. The slurry flows onto the pad and the carrier descends to contact the rotating pad. During rotation, the slurry is evenly distributed on the pad. The rotation of the platform creates relative motion between the wafer and the pad, generating shear forces that remove excess material from the wafer surface. The abrasive particles in the slurry remove material from the wafer, while the chemistry provides selectivity and optimizes polishing yield.

After the CMP process, the wafer surface is thoroughly cleaned to remove any residual slurry, particles or contaminants for analysis by post- process measurements.

Polishing experiments to determine material removal rates were conducted on 300 mm blanket wafers mounted on an Applied Materials 300 mm Reflexion polisher.

Polish removal experiments were conducted on 300 mm coated 15 kA-thick TEOS sheet wafers from Ramco, W coated wafers also from Ramco, and Ti and TiN coated wafers available from AMT Inc. All polishing experiments were conducted using H600 polyurethane polishing pads (commercially available from Fujibo Inc.) with the following, unless otherwise specified: typical down pressure 13.8 kPa (2.0 psi), CMP composition flow rate 300 mL/min, machine rotation speed 123 rpm, and carrier rotation speed 117 rpm. An A189L diamond pad conditioner (commercially available from 3M Company) was used to condition the polishing pads. The polishing pad and conditioner were run-in using 5.0 lb (2.3 kg) down force for 30 minutes at 101 rpm (platform) / 108 rpm (conditioner). Material removal rates for W were measured using a KLA-Tencor RS-100C metrology tool. Material removal rates for TEOS were measured using a KLA-Tencor OP-5300 metrology tool. Surface defect measurements using AFM :

Parksystem NX10 - Atomic force microscopy (AFM) is a technique used to image the surface of a sample with atomic resolution. Images are produced by scanning a sharp tip across the surface of the sample in a non-contact mode. The AFM probe is mounted on a cantilever, which acts as a tiny spring that detects the forces between the tip and the sample surface. These forces can include Van der Waals forces, electrostatic forces, magnetic forces, etc. The interaction between the tip and the sample results in tiny deflections of the cantilever.

The sample is placed on a stable stage under the AFM instrument, and the surface of the sample is clean and optimized for analysis. The AFM probe is brought close to the surface of the sample using a piezoelectric scanner. As the tip approaches the surface, the atomic forces between the tip and the sample become apparent. The piezoelectric scanner moves the sample stage to achieve the desired scan. The scanner records the height/position of the probe as it moves, using a laser beam deflection method to detect the deflection of the cantilever. A position-sensitive photodetector continuously monitors the deflection of the cantilever caused by atomic forces. A laser beam is directed to the back of the cantilever and the deflection of the cantilever is detected. This information is used in a feedback loop to adjust the vertical position of the probe to keep the deflection within the desired range. This maintains a constant force between the tip and the sample.

At each position, collect data on the height or displacement of the cantilever. Use this data to construct a topological image of the sample surface. Process and analyze the raw data to produce the final image. Use an algorithm to convert the height data into a visual representation. Sag measurement

Parksystem NX10 - Atomic force microscopy (AFM) is a technique used to image the surface of a sample with atomic resolution. Images are produced by scanning a sharp tip across the surface of the sample in a non-contact mode. The AFM probe is mounted on a cantilever, which acts as a tiny spring that detects the forces between the tip and the sample surface. These forces can include Van der Waals forces, electrostatic forces, magnetic forces, etc. The interaction between the tip and the sample results in tiny deflections of the cantilever.

The sample is placed on a stable stage under the AFM instrument, and the surface of the sample is clean and optimized for analysis. The AFM probe is brought close to the surface of the sample using a piezoelectric scanner. As the tip approaches the surface, the atomic forces between the tip and the sample become apparent. The piezoelectric scanner moves the sample stage to achieve the desired scan. The scanner records the height/position of the probe as it moves, using a laser beam deflection method to detect the deflection of the cantilever. A position-sensitive photodetector continuously monitors the deflection of the cantilever caused by atomic forces. A laser beam is directed to the back of the cantilever and the deflection of the cantilever is detected. This information is used in a feedback loop to adjust the vertical position of the probe to keep the deflection within the desired range. This maintains a constant force between the tip and the sample. At each position, data on the height or displacement of the cantilever is collected. This data is used to construct a topological image of the sample surface. The raw data is processed and analyzed to produce the final image. An algorithm is used to convert the height data into a visual representation.

The measurement is scanned from point A on the boundary to position B. For example, B is determined to be the center point in the silicon oxide/dielectric pillar/line, and its side is metal, such as tungsten (position A is located on tungsten). For both pre- and post-CMP samples, the step height is recorded as BA. The following presents the post-CMP minus pre-CMP height values, whereby there will be negative values in the case of recesses. Table 1 : Inventive Example - All concentrations are in wt.% relative to the total composition. Function Chemical Information IE1 IE2 IE3 IE4 IE5 IE6 Abrasive Silicon dioxide (surface modified) 6.5 5.0 7.15 6.5 6.5 6.5 Inhibitor 1 Lohexidine digluconate 0.02 0.02 0.02 0.005 0.02 0.02 Oxidants Fe(NO ₃ ) ₃ x 9H ₂ O 0.01 0.005 0.01 0.01 0.0075 0.01 Buffer* Arginine Arginine Arginine Arginine Arginine Arginine Stabilizer H4 _- EDTA 0.012 0.012 0.012 0.012 0.012 0.012 pH 4.0 4.0 4.0 4.0 4.0 3.0 * Unless otherwise specified, the buffer concentration is 0.5 wt.%. Table 1 continued. Function Chemical Information IE7 IE8 IE9 IE10 IE11 Abrasive Silicon dioxide (surface modified) 6.5 6.5 6.5 6.5 6.5 Inhibitor 1 Lohexidine digluconate 0.02 0.02 0.02 0.02 0.02 Oxidants Fe(NO ₃ ) ₃ x 9H ₂ O 0.005 0.01 0.005 0.01 0.01 Buffer* Arginine (0.3) Arginine (0.45) Arginine (0.7) Lysine Histidine Stabilizer H4 _- EDTA 0.012 0.012 0.012 0.012 0.012 pH Adjuster HNO ₃ 0.01 0.01 0.01 0.01 0.01 pH 4.0 4.0 4.0 4.0 4.0 * Unless otherwise specified, the concentration of buffer is 0.5 wt.%. Table 2 : Comparative Examples - All concentrations are in wt.% relative to the total composition. Function Chemical Information C1 C2 C3 C4 C5 C6 C7 C8 C9 Abrasive Silicon dioxide (surface modified) 6.5 6.5 3.0 6.5 6.5 6.5 6.5 6.5 Cationic Silica (surface modified) 2.0 Inhibitor 1 Lohexidine digluconate 0.02 0.02 0.02 - 0.02 0.02 0.02 0.02 - Inhibitor 2 Polyacrylamide 0.008 - - - - - - - 0.02 Oxidants Iron(III) nitrate 0.01 0.01 0.01 0.01 0.0025 0.005 0.01 0.005 0.01 Buffer* _{K2HPO4 ​} Glycine Arginine Arginine Arginine Arginine Arginine (0.8) K ₂ HPO ₄ - Stabilizer H4 _- EDTA 0.012 0.012 0.012 0.012 0.012 0.012 0.012 0.012 Malonic acid (0.07) pH 4.0 4.0 4.0 4.0 4.0 4.5 4.0 4.0 4.0 * Unless otherwise specified, the concentration of buffer is 0.5 wt.%. Experimental Results Table 3 Embodiment MRR TEOS [A/min] MRR W [A/min] SER W [ppb] TEOS/W MRR Ratio W Depression TEOS depression C1 556 91 11 6.1 13 -49 C2 559 118 19 4.7 -28 -35 C3 412 88 13 4.7 -51 -16 C4 554 153 498 3.6 -129 9 C5 557 69 8 8.1 -6 -104 C6 508 65 10 7.8 7 -51 C7 591 72 5 8.2 -14 -73 C8 569 110 13 5.2 -38 -16 IE1 545 98 10 5.6 -18 -19 IE2 473 75 7 6.3 -10 -38 IE3 664 113 11 5.9 -9 -29 IE4 569 104 14 5.5 -twenty one -14 IE5 539 91 9 5.9 -11 -20 IE6 556 110 9 5.1 -17 -twenty one IE7 513 88 12 5.8 -14 -19 IE8 535 95 15 5.6 -13 -twenty two IE9 588 75 6 7.8 -10 -30 IE10 539 93 9 5.8 -twenty one -twenty one IE11 554 89 11 6.2 -16 -twenty three Results Discussion

Table 3 shows the static etching rate (SER), material removal rate (MRR), and tungsten and dielectric (TEOS) recess values for different compositions. It is noted that for the inventive examples 1-11, the combination of various components (A) to (F) is critical in providing acceptable polishing efficiency and solution stability. The zeta potential of the unmodified silica particles was found to increase in an acidic system (pH < 7) (see Figure 1). However, surprisingly, the surface modified colloidal silica particles were found to constantly maintain a low zeta potential (range -35 to -60 mV), thereby providing colloidal stability, even in an acidic system. It was found that the substrate polished with the composition comprising cationic colloidal silica particles according to the present invention (FIG. 2a) exhibited a smooth surface, and the cationic colloidal silica particles (FIG. 2b) were found to produce agglomerates, which were clearly visible on the surface of the substrate (corresponding to Comparative Example C9). The addition of lohexidine or lohexidine digluconate as a corrosion inhibitor (B) plus a basic amino acid (D) having an isoelectric point (pI) ≥ 6.5 in the composition not only provided a SER of tungsten below 30 ppb in the provided pH range, but also provided suitably low recesses of both tungsten and dielectric (TEOS). It is further noted that the addition of lohexidine to the composition in the absence of component (D) (Comparative Example C1) or in the absence of component (B) (Comparative Example C4) resulted in undesirable effects such as high TEOS depression and high tungsten (W) SER and high W depression, respectively.

In addition, some basic amino acids with isoelectric points (pI) ≥ 6.5 were tested, and compositions containing these buffers showed suitable results, as can be seen from the results outlined in Table 3. On the other hand, it was noted that the presence of neutral amino acids such as glycine (Comparative Example C2) resulted in undesirably high tungsten pitting as well as TEOS pitting.

The compositions according to the embodiments of the claimed invention show the following improved properties: high _SiO2 (TEOS) MRR, low tungsten MRR, low tungsten SER, high tungsten suppression and optimal control of TEOS recess, while having colloidal or high dispersion stability and low alkali content.

without

The claimed invention is described in more detail by the attached drawings.

[Figure 1] shows the effect of pH on the zeta potential of surface modified colloidal silica particles (i.e. component (A)), measured by electrophoresis. Colloidal silica particles with two different particle sizes of 75 and 109 nm were studied.

[FIG. 2] depicts surface images (obtained by SEM) of substrates polished using compositions. FIG. 2a shows a substrate polished using a composition comprising cationic colloidal silica particles according to the present invention, while FIG. 2b shows a substrate polished using cationic colloidal silica particles.

[Figure 3] Depicts DLS measurement results of filtered silica dispersions measured with a Malvern Zetasizer ZSP when the zeta potential of the surface modified colloidal silica particles is > -35 mV at pH ranging from ≥ 2.0 to ≤ 4.5. The measurement results show that the DLS method cannot record stable measurements when the zeta potential of the surface modified colloidal silica particles is > -35 mV at pH ranging from ≥ 2.0 to ≤ 4.5. The zeta potential of the silica particles was set to -23 mV at pH 2.8 using 0.1 M KCl for measurement.

[Figure 4] Depicts DLS measurements of filtered silica dispersions measured with a Malvern Zetasizer ZSP when the zeta potential of the surface modified colloidal silica particles is < -35 mV at a pH range from ≥ 2.0 to ≤ 4.5.

Claims (14)

A dielectric polishing composition comprising: (A) surface-modified colloidal silica particles comprising negatively charged groups on the surface of the particles, wherein the surface-modified colloidal silica particles have a negative charge, a particle size of from 60 nm to 200 nm, and a zeta potential of ≤ -5 mV at a pH ranging from ≥ 2.0 to ≤ 6.0, wherein the concentration of the surface-modified colloidal silica particles (A) ranges from ≥ 3.2 wt.% to ≤ 13.0 wt.%; (B) at least one corrosion inhibitor selected from at least one guanidine derivative; (C) at least one iron (III) oxidant, wherein the concentration of the iron (III) oxidant (C) ranges from ≥ 0.003 wt.% to ≤ 0.1 wt.%; (D) at least one buffer, the buffer selected from at least one alkaline amino acid with an isoelectric point (pI) ≥ 6.5, wherein the concentration of the buffer (D) ranges from ≥ 0.1 wt.% to ≤ 0.78 wt.%; (E) at least one stabilizer, wherein the stabilizer (E) is selected from acetic acid, acetylacetonate, o-phosphoethanolamine, isophosphite, alendronic acid (alendronic acid) acid), acetic acid, phthalic acid, citric acid, adipic acid, oxalic acid, malonic acid, succinic acid, glutaric acid, pimelic acid, suppository acid, azelaic acid, sebacic acid, oxalic acid, citric acid, gluconic acid, muconic acid, ethylenediaminetetraacetic acid, propylenediaminetetraacetic acid, N,N-bis(carboxymethyl)alanine, nitrilotriacetic acid, diethylenetriaminepentaacetic acid, di(salicylidene)ethylenediamine, aminotris(methyleneisophosphite), diethylenetriaminepenta(methylisophosphite), ethylenediaminetetra(methyleneisophosphite) or a mixture thereof; and (F)   an aqueous medium, wherein the pH of the composition ranges from ≥ 2.0 to ≤ 4.3, and wherein the polyacrylamide content of the composition is ≤ 1 ppm. The composition of any one of claim 1, wherein the guanidine derivative is selected from buformin, phenformin, guanidine, proguanil hydrochloride, 2-guanidinebenzimidazole, polyhexamethylene biguanide hydrochloride, polyaminopropyl biguanide, alexidine, lohexidine or a lohexidine salt, preferably lohexidine or a lohexidine salt. The composition of any one of claims 1 to 2, wherein the pH of the composition ranges from ≥ 3.0 to ≤ 4.5. A composition as claimed in any one of claims 1 to 3, wherein the iron (III) oxidizing agent (C) is selected from iron (III) nitrate or a hydrate thereof. The composition of any one of claims 1 to 4, wherein the buffer (D) is a basic amino acid selected from lysine, arginine or histidine. A composition as claimed in any one of claims 1 to 5, wherein the composition has a potassium content of ≤ 1 ppm. The composition of any one of claims 1 to 6, wherein the composition further comprises an additive selected from a pH adjuster, an oxidant, a wetting agent, a dispersant, a biocide, or a mixture thereof. A composition as in any one of claims 1 to 7, wherein the composition is used for polishing a substrate (S), wherein the substrate (S) comprises: (i) tungsten and/or (ii) a tungsten alloy; and (iii) at least one dielectric layer selected from silicon, silicon oxide, silicon nitride or a low-k material. A process for manufacturing a semiconductor device, comprising chemically mechanically polishing a substrate (S) used in the semiconductor industry in the presence of a composition as defined in any one of claims 1 to 8, wherein the substrate (S) comprises (i)   tungsten and/or (ii)   tungsten alloy; and (iii)  at least one dielectric layer selected from silicon, silicon oxide, silicon nitride or low-k material. A process as claimed in claim 9, wherein a ratio of a material removal rate (MRR) of silicon oxide to a material removal rate (MRR) of tungsten ranges from 2:1 to 100:1. A process as in any one of claims 9 to 10, wherein the static etch rate (SER) of tungsten is less than 30 ppb. The process of any one of claims 9 to 11, wherein the material removal rate (MRR) of silicon oxide is > 300 Å/min. A process as claimed in any one of claims 9 to 12, wherein the material removal rate (MRR) of tungsten is less than 200 Å/min. A use of a composition as claimed in any one of claims 1 to 8 for polishing a substrate (S) comprising: (i) tungsten and/or (ii) a tungsten alloy; and (iii) at least one dielectric layer.