WO2011083475A1 - Sorbate-containing compositions for use in copper chemical mechanical planarization - Google Patents

Sorbate-containing compositions for use in copper chemical mechanical planarization Download PDF

Info

Publication number
WO2011083475A1
WO2011083475A1 PCT/IL2011/000019 IL2011000019W WO2011083475A1 WO 2011083475 A1 WO2011083475 A1 WO 2011083475A1 IL 2011000019 W IL2011000019 W IL 2011000019W WO 2011083475 A1 WO2011083475 A1 WO 2011083475A1
Authority
WO
WIPO (PCT)
Prior art keywords
copper
sorbate
concentration
composition
volume
Prior art date
Application number
PCT/IL2011/000019
Other languages
French (fr)
Inventor
Yair Ein-Eli
David Starosvetsky
Magi Margalit Nagar
Original Assignee
Technion Research & Development Foundation Ltd.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Technion Research & Development Foundation Ltd. filed Critical Technion Research & Development Foundation Ltd.
Publication of WO2011083475A1 publication Critical patent/WO2011083475A1/en

Links

Classifications

    • C—CHEMISTRY; METALLURGY
    • C09—DYES; PAINTS; POLISHES; NATURAL RESINS; ADHESIVES; COMPOSITIONS NOT OTHERWISE PROVIDED FOR; APPLICATIONS OF MATERIALS NOT OTHERWISE PROVIDED FOR
    • C09G—POLISHING COMPOSITIONS; SKI WAXES
    • C09G1/00—Polishing compositions
    • C09G1/02—Polishing compositions containing abrasives or grinding agents
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
    • H01L21/00—Processes or apparatus adapted for the manufacture or treatment of semiconductor or solid state devices or of parts thereof
    • H01L21/02—Manufacture or treatment of semiconductor devices or of parts thereof
    • H01L21/04—Manufacture or treatment of semiconductor devices or of parts thereof the devices having potential barriers, e.g. a PN junction, depletion layer or carrier concentration layer
    • H01L21/18—Manufacture or treatment of semiconductor devices or of parts thereof the devices having potential barriers, e.g. a PN junction, depletion layer or carrier concentration layer the devices having semiconductor bodies comprising elements of Group IV of the Periodic Table or AIIIBV compounds with or without impurities, e.g. doping materials
    • H01L21/30—Treatment of semiconductor bodies using processes or apparatus not provided for in groups H01L21/20 - H01L21/26
    • H01L21/31—Treatment of semiconductor bodies using processes or apparatus not provided for in groups H01L21/20 - H01L21/26 to form insulating layers thereon, e.g. for masking or by using photolithographic techniques; After treatment of these layers; Selection of materials for these layers
    • H01L21/3205—Deposition of non-insulating-, e.g. conductive- or resistive-, layers on insulating layers; After-treatment of these layers
    • H01L21/321—After treatment
    • H01L21/32115—Planarisation
    • H01L21/3212—Planarisation by chemical mechanical polishing [CMP]

Definitions

  • the present invention in some embodiments thereof, relates to chemical- mechanical planarization (CMP) and, more particularly, but not exclusively, to compositions that are useful for chemical-mechanical planarization (CMP) of copper- containing surfaces, to processes utilizing same and to products produced by these processes.
  • CMP chemical- mechanical planarization
  • Wafer-based electronic devices are manufactured with several levels of interconnected levels. As the number of interconnect levels increases, planarization of the previous layers is required to ensure a flat surface prior to subsequent lithography. Without planarization, the levels would become increasingly crooked and extend outside the depth of focus of available lithography, interfering with the ability to pattern. Chemical Mechanical Planarization (CMP) is the primary processing method to achieve such planarization.
  • CMP Chemical Mechanical Planarization
  • CMP is a process whereby chemical reactions increase the mechanical removal rate of a material, typically a metal.
  • CMP is commonly used to polish off high spots on wafers or other features deposited on wafers, flattening the feature or the wafer, referred to as planarization.
  • CMP is executed by use of slurries containing abrasive particles that effect the mechanical polishing, and reagents that effect the chemical reactions.
  • the chemical reaction that increases the mechanical removal rate is commonly tailored to provide a higher removal rate of one material versus another material.
  • the chemical action in CMP helps to achieve a higher removal selectivity of one material to another than a strictly mechanical process would provide.
  • Copper is widely used in the manufacturing of integrated circuits (IC) owing to its low electrical resistance (1.7 ⁇ /cm) and excellent electromigration resistance.
  • the integration of copper into IC is performed by a damascene technique, which involves CMP for removing the excess copper subsequent to wafer plating, in order to form a planar surface for subsequent fabrication of additional levels of chip wiring.
  • a high performance of the CMP process allows a successful fabrication of multilevel layers.
  • Typical copper CMP slurries consist of abrasives, such as silica particles, dispersed in an aqueous solution containing an oxidizer, such as hydrogen peroxide.
  • Hydrogen peroxide (H 2 0 2 ) is a powerful oxidizer and can be effectively used in various solutions in a wide range of pH values.
  • CMP slurries typically also require a complexing agent to dissolve the oxidized species and a corrosion inhibitor to form a passivating layer to protect the lower regions from excessive etching.
  • Glycine is an example of a widely used complexing agent, typically employed in acidic slurries. In peroxide-containing slurries, apart from glycine, few other complexing agents have also been proposed. These include, for example, acetic acid, ethylene diamine, glycolic acid, phthalic acid salt, citric acid, and oxalic acid.
  • Copper CMP can be performed in acidic or neutral or basic medium.
  • a high selectivity of Cu/Si0 2 polish rate is desirable to minimize erosion losses, as can be achieved with acidic CMP slurry.
  • acidic slurries are disadvantageous due to possible corrosion of the polishing equipment by the slurry.
  • ammonia and ammonium salt based slurries have been investigated but the reported Cu/Si0 2 polish rate selectivity was found to be an order of magnitude lower than that achievable in acidic slurries.
  • U.S. Patent Application No. 20070163677 teaches copper CMP slurry compositions which has a pH that ranges from 9 to 13, and which are devoid of ammonium-containing complexing agents.
  • dishing has been categorized as a parameter in CMP process that must be controlled. The dishing phenomenon was attributed to unbalanced copper passivation, uncontrolled dissolution of the copper structures and to the bending of the used CMP pad.
  • CMP slurry performance and its influence on dishing thus depend, at least in part, on the use of a corrosion inhibitor and on its characteristics.
  • the inhibitor is required to perform during three main stages; the copper CMP process itself, the over- polishing stage and the next stage down the line of the IC manufacturing process, which may employ different chemical/electrochemical media and different mechanical forces.
  • the inhibitor's ability to protect copper surface in such a dynamic process is determined by a few factors; the inhibitor's passivation/re-passivation kinetics, and the protective film stability in different pH regions and in the presence of oxidizing and complexing agents which are used throughout the IC manufacturing process.
  • Copper-corrosion inhibitors (also referred to herein and in the art as copper- etching inhibitors) have been described in the art. However, the integration of copper- corrosion inhibitors that perform well in solutions or other static media, in a dynamic process such as CMP is a non-trivial challenge.
  • a copper-corrosion inhibitor is employed in a CMP slurry in order to obtain a planarized and passivated surface by forming a protective film both at the recessed areas during the CMP process, and at the whole planarized surface once the CMP is completed (post-CMP).
  • a successful performance of a copper-etching inhibitor in a static medium therefore cannot predict a successful performance of the inhibitor is a dynamic process such as CMP.
  • BTA benzotriazole
  • copper corrosion inhibitors other than BTA.
  • Other additives proposed as copper corrosion inhibitors include, for example, citric acid, hydrazine, amino tetrazole, potassium sorbate [Abelev, E. et al, Electrochim. Acta 52 (2007), p. 1975], and imidazole.
  • Other agents used for passivating copper are taught in, for example, U.S. Patent Nos. 5,770,095 and 6,569,350 and U.S. Patent Application No. 20070163677.
  • Potassium sorbate (potassium (2E,4E)-hexa-2,4-dienoate, CAS number 24634- 61-5, 150.22 grams per mol, hereinafter K-sorbate) is the potassium salt of sorbic acid, a naturally occurring carboxylic acid, which finds its primary use in the food and personal care industries as the preservative ⁇ 202. Sorbate salts are generally accepted as safe and non-toxic, and are typically very soluble in water, up to 58.2 % at 20 °C in the case of potassium sorbate. Being available at relatively low cost and being safe, sorbate salts such as potassium sorbate, are highly desirable alternative reagents in many industries.
  • the present invention in some embodiments thereof, relates to highly effective chemical-mechanical planarization (CMP) of copper using sorbate as a copper passivation agent in high concentration of more than 0.2 molar.
  • CMP chemical-mechanical planarization
  • These CMP compositions surpass known compositions in many criteria in terms of process and end- product characteristics, such as copper removal rate, dishing, smoothness, and post- CMP conditioning of the planaraized copper surface.
  • a copper CMP composition that includes hydrogen peroxide, glycine and a sorbate anion, wherein the concentration of sorbate in the composition is at least 0.2 mol per liter.
  • a copper CMP composition that includes:
  • a complexing agent in a concentration that ranges from 1 gram per liter to 40 grams per liter;
  • the concentration of the peroxide ranges from 5 percent by volume to 10 percents by volume.
  • the peroxide is hydrogen peroxide.
  • the concentration of the complexing agent ranges from 5 grams per liter to 30 grams per liter.
  • the complexing agent is selected from the group consisting of an amino acid, a monoamine-containing compound and a diamine- containing compound.
  • the complexing agent is glycine.
  • the concentration of sorbate is at least 1.7 mol per liter.
  • the source of the sorbate anion is potassium sorbate.
  • the concentration of hydrogen peroxide ranges from 0.1 percent by volume to 10 percents by volume, or from 5 percent by volume to 10 percents by volume, or from 0.3 percent by volume to 1 percent by volume.
  • the concentration of glycine ranges from 5 grams per liter to 40 grams per liter, or from 5 grams per liter to 30 grams per liter, or from 10 grams per liter to 30 grams per liter.
  • the concentration of the sorbate anion ranges from 0.4 mol per liter to 9 mol per liter, a concentration of glycine is 20 grams per liter, and a concentration of hydrogen peroxide ranges from 5 percent by volume to 10 percents by volume.
  • the composition presented herein has a pH that ranges from 7 to 10, or a pH of 8.5.
  • composition presented herein further includes abrasive particles.
  • the abrasive particles are selected from the group consisting of silica, alumina, titania, zirconia, ceria, germanium oxide and mixtures thereof.
  • a process of chemical-mechanical planarization of a copper-containing surface which is effected by contacting a substrate having a copper-containing surface with the copper CMP slurry composition presented herein, thereby planarizing the copper-containing surface.
  • the process includes, subsequent to the contacting, applying a polishing pad onto the substrate.
  • the process is characterized by a copper removal rate of at least 200 nanometer per minute. In some embodiments, the process is characterized by a planarized copper- containing surface that has a copper dishing value of less than 110 nanometer.
  • a substrate having a planarized copper-containing surface obtained by the process presented herein.
  • the planarized copper-containing surface is characterized by a height range (Z range) value of less than 20 nm.
  • the planarized copper-containing surface is characterized by a water-to-surface contact angle of at least 80 °.
  • the substrate forma a part of an article that comprises a copper-containing element.
  • the article is selected from the group consisting of an integrated circuit, a device wafer, a 3D integration device and an advanced interconnect device.
  • an article that includes a wafer-based device having at least one copper-containing element, the copper-containing element being produced by the process presented herein.
  • the article is characterized by a copper dishing value of less than 110 nanometers and/or a height range (Z range) value of less than 20 nm.
  • FIG. 1 is a comparative plot showing two anodic polarization curves of copper obtained by polarizing a copper electrode in a solution containing one gram per liter (g/L) of potassium sulfate (1 g/L K 2 SO 4 , curve denoted 'a 1 ), and from polarizing copper in a solution containing one gram per liter of potassium sulfate and 8 grams per liter of potassium sorbate (curve denoted b 1 );
  • FIG. 2 is a comparative plot showing three anodic polarization curves of copper obtained by polarizing a copper electrode in a solution containing 4 g/L potassium sorbate (K-sorbate), each further containing either 1 g/L of potassium sulfate (K- sulfate), 1 g/L glycine, or 1 g/L potassium acetate (K-acetate);
  • FIG. 3 is a comparative plot showing four anodic polarization curves of copper obtained by polarizing a copper electrode in a solution containing 1 g/L Na 2 S0 4 and 10 g/L glycine, and containing no hydrogen peroxide (curve denoted ⁇ 1 ), 0.3 percents by volume hydrogen peroxide (curve denoted '2"), 1 percents by volume hydrogen peroxide (curve denoted '3 and 5 percents by volume hydrogen peroxide (curve denoted '4');
  • FIGs. 4A-F present high resolution SEM micrographs (FIGs. 4A, 4C and 4E, left column) alongside with plots presenting potentiostatic measurements (FIGs. 4B, 4D and 4F respectively, right column), obtained from a copper electrode polarized at 0.3 (scE) for 200 seconds in a solution containing 1 g/L Na 2 S0 4 andlOg/L glycine and containing 0.3 percents by volume hydrogen peroxide (FIGs. 4A-B), 1 percents by volume hydrogen peroxide (FIGs. 4C-D), or 5 percents by volume hydrogen peroxide (FIGs. 4E-F); FIGs.
  • FIGs. 5A-F present comparative plots of potentiodynamic curves (FIGs. 5A, 5C and 5E, left column) alongside with comparative plots of potentiostatic curves at 300 mV (FIGs. 5B, 5D and 5F respectively, right column), obtained from a copper electrode polarized in solutions containing 1 g/L Na 2 S0 4 , 10 g/L glycine and containing 0.3 percents by volume hydrogen peroxide (FIGs. 5A-B), 1 percents by volume hydrogen peroxide (FIGs. 5C-D), or 5 percents by volume hydrogen peroxide (FIGs. 5E-F), and further containing different K-sorbate concentrations as indicated in each plot;
  • FIG. 6 presents a comparative plot showing the copper removal rate as a function of sorbate content in slurries containing 10 g/L glycine with low or intermediate concentrations of H 2 0 2 ;
  • FIG. 7 presents a comparative plot of three copper removal rate curves using CMP slurries containing three different H 2 0 2 concentrations and containing K-sorbate at a concentration of 16 g L, each curve plotted as a function of glycine concentration and 3 different H 2 0 2 concentrations;
  • FIGs. 8A-D presents comparative plots of high-resolution X-ray photoelectron spectra (XPS) of Cu2p 3/2 obtained from copper surface subsequent to CMP process in slurries containing 64 g/L K-sorbate, 10 g/L glycine and three different H 2 0 2 concentrations (0.1, 1 and 10 percent by volume presented in FIG. 8C, FIG. 8B and FIG. 8A respectively) and recorded at a take-off angle of 12° relative to the copper surface, whereas etched copper sample spectrum serves as a reference (FIG. 8D);
  • XPS X-ray photoelectron spectra
  • FIGs. 9A-C presents comparative plots of high-resolution X-ray photoelectron spectra (XPS) of Ols obtained from copper surface subsequent to CMP process in slurries containing 64 g/L K-sorbate, 10 g/L glycine and different H 2 0 2 concentrations (0.1, 1 and 10 percent by volume presented in FIG. 9C, FIG. 9B and FIG. 9A respectively);
  • XPS X-ray photoelectron spectra
  • FIGs. 10A-B present Cu2p 3 2 (FIG. 10A) and Ols (FIG. 10B) X-ray photoelectron spectra (XPS) at different take-off angles taken from the surface of a copper sample subsequent to CMP process in slurry containing 64 g/L K-sorbate, 10 g/L glycine and low H 2 0 2 concentration of 0.1 percent by volume.
  • the take-off angle of the bulk mode (69° related to the copper surface) provides maximum information from the bulk;
  • FIGs. 11A-B present Cu2p 3/2 (FIG. 11A) and Ols (FIG.
  • XPS X-ray photoelectron spectra
  • FIG. 12 presents comparative plots of extrapolated data estimated from the intensity ratio (Io/I m ) between the oxide and the metal peaks obtained in XPS measurements, showing the oxide thickness developed on copper as a function of H 2 0 2 content in the slurry having 64 g/L K-sorbate and 10 g/L glycine;
  • FIGs. 13A-D present contact angle measurements of a water droplet at a copper wafer surface (FIGs. 13A and 13B) and SEM images (FIGs. 13C and 13D) obtained from the copper wafer surface, subsequent to CMP process in slurries containing 64 g/L K-sorbate and 10 g/L glycine with H 2 0 2 concentrations of 0.1 percent by volume (FIGs. 13 A and 13C) and of 5 percents by volume (FIGs. 13B and 13D);
  • FIG. 14 presents comparative plots showing the XPS intensities ratio of Cu(OH) 2 and CuO, normalized to the Cu° peak in the Cu2p 3/2 spectrum, as measured at an angle of 12 ° related to the copper surface subsequent to polishing with slurries containing 64 g/L K-sorbate, 10 g/L glycine and different H 2 0 2 concentrations;
  • FIGs. 15A-D present contact angle measurements of a water droplet at a copper wafer surface (FIGs. 15A and 15B) and SEM images (FIGs. 15C and 15D) obtained from the copper wafer surface, subsequent to CMP process in slurries containing 10 g/L glycine and 0.1 percent by volume 3 ⁇ 40 2 with sorbate concentrations of 4 g/L (FIGs. 15 A and 15C) and of 64 g/L (FIGs. 15B and 15D);
  • FIGs. 16A-D summarizes the postulated processes occurring on the surface of copper when using the H 2 0 2 -glycine slurries containing sorbate, as concluded from the experimental results discussed hereinabove.
  • FIGs. 16A-D are schematic diagrams depicting a theoretical surface of a copper sample polished by slurries having two concentrations of H 2 0 2 , 0.1-0.3 percent by volume (low H 2 0 2 , FIGs. 16 A and 16B) and 5-10 percent by volume (high H 2 0 2 , FIGs. 16C and 16D), and two sorbate concentrations, 4g/L (low sorbate, FIGs. 16A and 16C) and 64 g/L (high sorbate, FIGs. 16B and 16D);
  • FIG. 17 presents a 4X5 array of 20 optical microscope images of copper surface subsequent to CMP process with CMP slurries containing different concentration of sorbate and H 2 0 2 and a fixed glycine concentration of 10 g/L, denoting the removal rate (RR in nanometer per minute, nm/min) values, wherein the concentration of H 2 0 2 increases column-wise 0.1, 0.3, 1, 5 and 10 percents by volume, and the concentration of sorbate increases row-wise 0, 4, 16 and 64 g/L;
  • FIGs. 18A-B present AFM micrographs of two copper samples subsequent to polishing with slurries having 4 g L (FIG. 18A) and 64 g/L (FIG. 18B) sorbate, wherein the roughness of the sample treated with 4 g/L sorbate exhibited a Z range of 37.319 nm, an RMS (Rq) of 1.621 nm, and a mean roughness (Ra) of 1.306 nm, and the sample treated with 64 g/L sorbate exhibited a Z range of 13.533 nm, an RMS (Rq) of 1.284 nm, and a mean roughness (Ra) of 0.99 nm;
  • FIGs. 19A-D present optical microscope images of copper bond pads (100 x 100 ⁇ 2 ) subsequent to copper and barrier CMP, using copper CMP slurries having 4 g/L sorbate (FIGs. 19A and 19C) or 64 g/L sorbate (FIGs. 19B and 19D), polished by an edge die (FIGs. 19A and 19B) or a center die (FIGs. 19C and 19D), demonstrating the surface finish and polishing uniformity as a function of the sorbate concentration;
  • FIGs. 20A-D present SEM images of 90 nm copper wiring, showing top view (FIGs. 20A and 20B) and side view (FIGs. 20C and 20D), polished with CMP slurries having 4 g/L sorbate (FIGs. 20A and 20C) or 64 g/L sorbate (FIGs. 20B and 20D), wherein copper corrosion sites are marked by arrows in FIG. 20C;
  • FIG. 21 presents the dishing values of copper van der-Pauw (VDP) structures
  • FIG. 22 presents a box plot of normalized copper sheet resistance values at different copper densities (PD) as a function of sorbate concentrations, wherein the sheet resistance values are normalized to a maximum value obtained in sample that was polished with a slurry having a sorbate concentration of 4 g/L;
  • PD copper densities
  • FIG. 23 presents a box plot of normalized copper sheet resistances values as a function of sorbate concentration in the slurry measured from different line widths, wherein the sheet resistance values are normalized to a maximum value obtained from sample polished with a slurry having a sorbate concentration of 4 g/L;
  • FIG. 24 summarizes the CMP process with respect to sorbate concentration in a
  • FIGs. 24A-D are schematic diagrams depicting a theoretical surface of a copper sample polished by slurries having two sorbate concentrations, 4 g/L (low sorbate, FIGs. 24A and 24C) and 64 g L (high sorbate, FIGs. 24B and 24D), in two polishing stages, the main CMP process (FIGs. 24A and 24B) and the over-polishing stage (FIGs. 24C and 24D); and
  • FIGs. 25A-D are electron micrographs at a magnification of 8,000X (FIG. 25A),
  • FIG. 25B 64,000X (FIG. 25B), ⁇ , ⁇ (FIG. 25C) and 50,000X (FIG. 25D) of copper- containing wafer samples subjected to CMP before post-treatment with 0.4 M sorbate (FIGs. 25A and 25B) and after post-treatment with 0.4 M sorbate (FIGs. 25C and 25D), showing significant decrease in the silica particles adhesiveness as a result of the surface passivation effect imparted by sorbate to the copper surface.
  • the present invention in some embodiments thereof, relates to chemical- mechanical planarization (CMP) and, more particularly, but not exclusively, to compositions that are useful for chemical-mechanical planarization (CMP) of copper- containing surfaces, to processes utilizing same and to products produced by these processes.
  • CMP chemical- mechanical planarization
  • CMP chemical mechanical planarization
  • recursive dynamically re-occurring
  • oxidation mechanical removal
  • abrasive chemical removal
  • complexing chemical re-passivation
  • corrosion inhibition rapid chemical re-passivation
  • a successful CMP process depends, at least in part, on the chemical reagents used in the process, or, in other words, on the chemical reagents composing a CMP slurry composition.
  • the corrosion and passivation of copper during CMP is substantially different than static corrosion or static passivation of copper, since the latter do not involve these reactions in a concurrent situation, and further do not occur with recursive mechanical removal of surface layers.
  • a successful CMP slurry composition should perform so as to balance between removal of copper and passivation of the exposed copper surface. Such a balance is achieved by selecting an oxidizing agent, a complexing agent and a copper-corrosion inhibitor that together perform so as to afford a chemically stable (passivated) and smoothly planarized copper surface.
  • Prasad, Y.N. et al. [Electrochimica Acta, 52 (2007), pp. 6353-6358] have studied the effect of various copper-etching inhibitors on the static etch rate of copper, when added to a CMP slurry composition containing hydrogen peroxide as an oxidizer (typically at a concentration of 5 weight percents) and arginine as an alkaline complexing agent.
  • a CMP slurry composition containing hydrogen peroxide as an oxidizer (typically at a concentration of 5 weight percents) and arginine as an alkaline complexing agent.
  • the present inventors have recognized that an effective performance of sorbate in the dynamic CMP process depends, at least in part, on the sorbate concentration, as well as on type and/or concentration of the other components in the CMP composition, as is further supported by the above-described teachings.
  • CMP composition such as the oxidizing agent and the complexing agent, and have outlined some features that are required for an optimal and well-balanced copper CMP slurry composition which affords simultaneously a high copper removal rate and a rapid and stable re-passivated copper surface during and after the CMP process.
  • a copper CMP composition which includes:
  • a complexing agent in a concentration that ranges from 1 gram per liter to 40 grams per liter;
  • CMP composition describes a composition useful in a chemical-mechanical planarization process, as this term is widely understood in the art.
  • a CMP composition is thus comprised from abrasive particles for effecting mechanical planarization and from chemical reagents for effecting chemical processes. Accordingly, a CMP composition is often in a form of a slurry and is referred to herein and in the art also as “CMP slurry composition” or simply as “CMP slurry”.
  • copper CMP composition describes a CMP composition that is suitable for chemical-mechanical planarization of copper-containing surfaces.
  • copper-containing surfaces it is meant a substance that has a surface, and wherein the surface comprises at least 50 weight percents copper, at least 60 weight percents copper, at least 70 weight percents copper, at least 80 weight percents copper, at least 90 weight percents copper, at least 95 weight percents copper, or being 100 percents copper.
  • a copper CMP composition typically includes an oxidizer that is suitable for effecting copper corrosion and a complexing agent that is suitable for complexing copper ions in the formed copper oxides, and thus for dissolving copper oxides.
  • the substance may be made, at least in part, from copper, or can include copper only at its surface, as a coating.
  • the copper can be found on the surface of the substrate in any shape and form, such as dots, lines, polygonal or circular shaped areas and any combination thereof.
  • “molar”, also denoted as "M” refers to an absolute concentration of the described substance in the CMP composition, namely, by mols per liter composition.
  • grams per liter it is meant a weight of the substance per 1 liter of the composition.
  • the CMP slurry composition is essentially aqueous, as the process is effected in a conducting medium such as water.
  • the concentration of the sorbate anion can be expressed in grams per liter potassium sorbate, hence concentrations higher than 0.2 mols per liter correspond to concentrations of potassium sorbate (150.22 grams per mol) higher than about 30 grams/L.
  • sorbate anion may include, without limitation, sodium sorbate and other soluble salts of sorbic acid.
  • the role of the sorbate seems to negate the role of the other two components in the CMP composition (the oxidizing agent and the complexing agent), as the sorbate associates with both freshly exposed copper and freshly formed copper oxides surface and protects the copper from being further corroded (passivation and re- passivation role), while the oxidizing agent oxidizes the copper to various copper- oxygen species which are more soluble and less resistant to mechanical polishing, and copper complexing agent assists the copper removal process by tipping the equilibrium further towards copper dissolution.
  • Example 2 when following the parameter of copper removal rate from blanket wafers, optimal performance of copper planarization and passivation is obtained at sorbate concentrations higher than 0.2 M sorbate (30 grams/L K-sorbate) under conditions of high concentrations of H 2 0 2 , indicating that the corrosive effect of the oxidizer has reached saturation, and the protective effect of the sorbate becomes apparent by lowering the rate of copper removal (see, for example, Table 2 and Table 3 hereinbelow).
  • CMP compositions that contain from 16 grams per liter (about 0.1 M sorbate anion) to 64 grams per liter (about 0.4 M sorbate anion) perform better than compositions containing 4 grams per liter (about 0.025 M sorbate anion).
  • XPS measurements follow the chemical composition of the sorbate-passivated copper-containing substrate surface at various angles (corresponding to probing the surface at various depths), contact angle measurements quantify the change in the surface tension of a copper-containing substrate as a result of passivation with sorbate, and AFM measurements quantify the smoothness of the planarized copper surface as a function of the sorbate content in the CMP slurry.
  • the sorbate anion concentration in the copper CMP slurry is higher than 0.2 M, higher than 0.4 M, higher than 0.5 M, higher than 1.7 M or higher than 2 M.
  • its concentration in the copper CMP slurry is higher than 30 grams/L, higher than 40 grams/L, higher than 50 grams/L. higher than 60 grams/L, higher than 64 grams/L, higher than 70 grams/L, higher than 80 grams/L, higher than 90 grams/1, higher than 100 grams/L, and even higher than 200 grams/1 or higher than 256 grams/L.
  • a concentration of a sorbate anion in the CMP slurry can be, for example, 0.200 M, 0.206 M, 0.213 M, 0.220 M, 0.226 M, 0.233 M, 0.240 M, 0.246 M, 0.253 M, 0.260 M, 0.266 M, 0.273 M, 0.280 M, 0.286 M, 0.293 M, 0.300 M, 0.306 M, 0.313 M, 0.320 M, 0.326 M, 0.333 M, 0.340 M, 0.346 M, 0.353 M, 0.359 M, 0.366 M, 0.373 M, 0.379 M, 0.386 M, 0.393 M, 0.399 M, 0.406 M, 0.413 M, 0.419 M, 0.426 M, 0.433 M, 0.439 M, 0.446 M, 0.453 M, 0.459 M, 0.466 M, 0.473 M, 0.479 M, 0.486 M, 0.493 M, 0.499 M, 0.506 M
  • sorbate is characterized by having an exceptional high solubility in water (9 M), which allows using high concentrations in a CMP aqueous composition. It is further noted that as discussed herein and is further demonstrated in the Examples section that follows, such high concentrations of sorbate are required to achieve the balance between rapid corrosion versus rapid re-passivation in a dynamic CMP process, and further achieve desired characteristics of the obtained planarized copper surface, probably due to the sorbate rapid adsorption to copper.
  • balancing the requirements for a high copper removal rate and rapid re-passivation of the freshly polished copper surface require balancing the rates of copper passivation, effected by the sorbate anion, versus copper corrosion, effected by the oxidizing agent and the complexing agent.
  • an oxidizing agent such as hydrogen peroxide has been identified as providing a layer of copper oxides/hydroxides species at varying rates, which depend on the H 2 0 2 concentration in the CMP composition (see, the Examples section that follows).
  • H 2 0 2 concentration in the CMP composition see, the Examples section that follows.
  • the corrosive reaction which is required for the CMP process can be achieved by various oxidizing agents other than a peroxide.
  • a composition of the present invention may include any type of copper oxidizing agents, including, for example, peroxides, phenols, permanganates, chromates, iodates, iron salts, aluminum salts, sodium salts, potassium salts, phosphonium salts, chlorates, perchlorates, persulfates oxidizing agents, and any mixtures thereof.
  • copper oxidizing agents including, for example, peroxides, phenols, permanganates, chromates, iodates, iron salts, aluminum salts, sodium salts, potassium salts, phosphonium salts, chlorates, perchlorates, persulfates oxidizing agents, and any mixtures thereof.
  • Suitable oxidizing agents include, but are not limited to, phenol,
  • a peroxide is used as an oxidizing agent capable of oxidizing copper to copper oxide/dioxide.
  • the term "peroxide” is a chemical substance containing the peroxide anion (0 2 ⁇ ) or a single oxygen-oxygen bond, and capable of forming, even as an intermediate, a reactive oxygen radical species.
  • Peroxides include both inorganic peroxides and organic peroxides.
  • Non-limiting examples of inorganic peroxides include hydrogen peroxide, sodium peroxide, barium peroxide, calcium peroxide and magnesium peroxide.
  • Organic peroxides include, without limitation, benzoyl peroxide and methyl ethyl ketone peroxide.
  • the peroxide is used in a concentration that ranges from 5 percent by volume to
  • the peroxide is hydrogen peroxide (H 2 0 2 ).
  • the complexing agent pushes the equilibrium towards copper dissolution by forming complexes with various copper ions which form as a result of the oxidation reaction.
  • the complexing agent can be any copper complexing agent known in the art which is suitable for use in neutral and alkaline (basic) slurries.
  • a suitable copper complexing agent is such that, in combination with the high sorbate concentrations disclosed herein and the peroxide oxidizing agent, performs such that the required balance between copper removal rate (copper dissolution) and copper passivation is achieved under dynamic CMP conditions.
  • a suitable copper complexing agent is selected such that the increase in the corrosion inhibition activity of sorbate exhibited at increased sorbate concentrations is not reversed or flattened in the presence of the complexing agent, and vice versa, that the inhibitory effect of sorbate does not diminish the effective copper dissolution enhancement activity of the complexing agent.
  • Non-suitable complexing agents are those with which the above-described performance of sorbate is reversed or flattened, and/or those that cannot effect copper complexation effectively in high sorbate concentrations.
  • suitable complexing agents include, without limitation, amine-containing agents, such as, but not limited to, amino acids, monoamine-containing compounds and diamine-containing compounds.
  • amine-containing agents such as, but not limited to, amino acids, monoamine-containing compounds and diamine-containing compounds.
  • Exemplary diamines with aliphatic linear carbon chain include compounds such as ethylene diamine (1,2-diaminoethane), 1,3-diaminopropane (propane-l,3-diamine), putrescine (butane- 1,4-diamine), cadaverine (pentane-l,5-diamine) and hexamethylenediamine (hexane-l,6-diamine).
  • exemplary diamines include ethambutol, dimethyl-4-phenylenediamine, N,N'-di-2-butyl-l,4-phenylenediamine, diphenylethylenediamine and 1,8-diaminonaphthalene.
  • exemplary suitable complexing agents include glycolic acid, phthalic acid salt, citric acid, and oxalic acid.
  • Amino acids which are suitable complexing agents include, but are not limited to, glycine, arginine, phenyl alanine and serine.
  • a performance of a CMP composition which contains a high concentration of sorbate can be attenuated by the concentration of the complexing agent used.
  • the concentration of the complexing agent in the CMP composition ranges from 0.07 mol per liter (M) to 0.5 mol per liter (M), or from 0.1 M to 0.5 M.
  • the concentration of the complexing agent in the CMP composition is about 0.07 M, about 0.1 M, about 0.2 M, about 0.3 M, about 0.5 M, or about 0.5 M.
  • Prasad, Y.N. et al. used arginine as a complexing agent with sorbate, but have failed to show a beneficial effect of combining the two agents, presumably since Prasad, Y.N. et al. have used non-dynamic criteria to assess their compositions; hence, showing that the etch rate increased with an increase of sorbate concentration while studying arginine-containing slurry compositions, probably due to changes in the electrical conductance.
  • the complexing agent is glycine.
  • a copper CMP composition which includes hydrogen peroxide as an oxidizing agent, glycine as a complexing agent and a sorbate anion at a concentration higher than 0.2 mol per liter.
  • the concentration of hydrogen peroxide in the copper CMP slurry composition may range from 0.1 percent by volume to 10 percents by volume.
  • the concentration of H 2 0 2 may range from 0.3 percent by volume to 1 percents by volume (referred to as low H 2 0 2 concentration), and further alternatively, the concentration of H 2 0 2 ranges from 5 percent by volume to 10 percents by volume (referred to as high H 2 0 2 concentration) (see, e.g., Example 5 hereinbelow).
  • the concentration of glycine in the copper CMP slurry composition may range from 5 grams per liter to 40 grams per liter.
  • the concentration of glycine ranges from 5 grams per liter to 10 grams per liter (see, e.g., Example 2 hereinbelow).
  • CMP slurry composition includes sorbate anion in a concentration that ranges from 0.2 mols per liter to 9 mols per liter, glycine in a concentration of 10 grams per liter (about 0.07 M), and hydrogen peroxide in a concentration that ranges from 5 percent by volume to 10 percents by volume.
  • the pH of the copper CMP composition ranges from 7 to 14, optionally, from 7 to 10, optionally from 8 to 9, and optionally the pH is about 8.5.
  • the pH can be set to the desired level by adding small amounts of base or acid.
  • the pH of the otherwise acidic solution of hydrogen peroxide, sorbate and glycine can be adjusted to 8.5 by adding KOH, NaOH or other suitable bases.
  • a CMP composition described herein can further include a pH-adjusting agent (e.g., a base, such as an alkali metal hydroxide, for example, KOH or NaOH).
  • a pH-adjusting agent e.g., a base, such as an alkali metal hydroxide, for example, KOH or NaOH.
  • a suitable complexing agent in the context of embodiments of the present invention is one that is suitable for alkaline slurry compositions, such as amine-containing complexing agents.
  • a copper CMP composition as described herein may further include additional components, including, for example, additional complexing agents, and/or metal salts (e.g., potassium sulfate.
  • the composition presented herein further comprises abrasive particles.
  • the abrasive particles are used to effect mechanical planarization and are selected to selectively remove passivated copper (in the form of copper oxides).
  • Exemplary particles are made of metal oxides.
  • Suitable metal oxides include, but are no limited to, oxides of silicon, aluminum, cerium, germanium, titanium, zirconium and mixtures thereof.
  • Specific examples of suitable abrasives include particles of Si0 2 , Ce0 2 , A1 2 0 3 , SiC, Si 3 N 4 and Fe 2 0 3 .
  • the abrasive particles make up between about 1 % and 30 % by weight of the composition.
  • a process of chemical-mechanical planarization of a copper-containing surface which is effected by contacting a substrate having a copper-containing surface with the copper CMP slurry composition as presented herein, thereby planarizing the copper-containing surface.
  • This process is also referred to herein as CMP process.
  • a CMP process involves a simultaneous chemical (e.g. by copper-oxidizing and copper-complexing agents) and mechanical (by means of abrasives) planarization of copper-containing surfaces, and a well balanced CMP process also provides a freshly exposed yet passivated copper surface. It has also been discussed hereinabove, that planarization of the bases layers in a multi-level wafer- based device manufacturing is required to ensure a flat surface prior to subsequent layering and/or lithography, and that passivation is required to maintain the planarization though the various steps of the process. Hence, a beneficial CMP process is required to produce a completely flat and smooth surface of the base layer that stays protected from further uncontrolled corrosion.
  • planarizing a copper-containing surface is defined herein as the removal of protruding or uneven parts of the surface in a copper-containing substrate so as to arrive at a planar and leveled surface at a minimal loss of substance.
  • the CMP process would also provide, apart from planarization, a passivated copper surface, namely planarized copper surface which would not undergo further etching or uncontrolled corrosion.
  • the process further includes, subsequent to contacting the substrate with the slurry, applying a polishing pad onto the substrate to effect polishing of the copper-containing surface of the substrate.
  • the process is characterized by a copper removal rate of at least 200 nanometer per minute under a given set of mechanical conditions, such as those presented for example in Table 1 in Example 1 hereinbelow. It is noted that optimal removal rates would be different using identical slurries under different mechanical conditions; hence the rate of copper removal should be regarded as "normalized” against any given set of mechanical conditions.
  • dishing refers to the phenomenon of uneven, non- planar surface polishing result, wherein sunk regions are formed typically near the center of an otherwise flat/planar area due to various reasons.
  • dishing is associated with uncontrolled corrosion of the surface material, and can be limited or resolved by rapid re-passivation of the surface during the CMP process, which involved dynamic and simultaneous corrosion and re-passivation reactions.
  • dishing also depends on the span and geometry of the polished surface, as small areas will be less prone to dishing compared to wide areas.
  • parameterization of dishing should refer to mechanical parameters, as in the case of removal rates discussed hereinabove, and to the size and shape of the copper surface being planarized.
  • the process is characterized by a dishing value of less than 110 nanometer (see, e.g., Example 8 hereinbelow) when measured under a given set of mechanical conditions as presented for example in Table 1 in Example 1 hereinbelow, and given a sample of copper having a van der-Pauw (VDP) structure of 80 ⁇ x 80 ⁇ in size.
  • a dishing value of less than 110 nanometer (see, e.g., Example 8 hereinbelow) when measured under a given set of mechanical conditions as presented for example in Table 1 in Example 1 hereinbelow, and given a sample of copper having a van der-Pauw (VDP) structure of 80 ⁇ x 80 ⁇ in size.
  • VDP van der-Pauw
  • the intermediate product of a manufacturing process involving copper CMP is typically a substrate having a planarized and passivated copper-containing surface
  • a substrate having a planarized and passivated copper-containing surface obtained by the process presented herein, using the CMP slurry composition presented herein.
  • While the end-product of manufacturing a wafer-based electronic device may have copper-containing surfaces embedded therein which have been processed by the presently disclosed CMP compositions and processes, these surfaces may be inaccessible due to layers that have been added on top during the formation of a complete electronic device.
  • the quality of the entire process of manufacturing electronic wafer-based devices which is also measured by the rate of disqualification of malfunctioning devices on a wafer, is however improved by using the compositions and processes disclosed herein.
  • the planarized and passivated copper-containing surface may be characterized by height range (Z range representing roughness/smoothness) value of less than 20 nm or less than 15 nm (see, e.g., Example 6 hereinbelow).
  • the presence of a protective (passivating layer) of sorbate on the planarized and passivated copper- containing surface can be determined and even quantified by the shape of a drop of deionized water placed on the planarized and passivated copper-containing surface.
  • the copper surface may by further characterized by a water-to-surface contact angle of at least 80 ° (see, e.g., Example 4 hereinbelow).
  • a substrate having a planarized copper-containing surface, obtained by the composition and process presented herein can be characterized, is the resistance of a sheet of copper processed thereby.
  • Such copper sheets or patterned copper lines with various line widths can be used to simulate an integrated wafer-based circuit for purposes of evaluating the performance of the CMP process and compositions.
  • sheet resistance values were obtained from samples treated by the process and compositions presented herein, and normalized against the values obtained from an identical sample processed with a composition containing a low concentration of 0.03 M sorbate (corresponding to about 4 g L K-sorbate).
  • the copper sheet resistance values characterizing a substrate having a planarized copper-containing surface treated by the process and compositions presented herein, and having copper pattern density 10-60 % ranges from 0.2 arbitrary units (A.U.) to 0.4 A.U.
  • the copper sheet resistance values characterizing a substrate having a planarized copper-containing surface treated by the process and compositions presented herein, and having copper lines width of 1-30 ⁇ ranges from 0.1 A.U. to 0.3 A.U.
  • Another criterion for assessing the benefits of a CMP composition is the "cleanness" of the copper surface at the end of the CMP process, namely the absence of remaining abrasive particles thereon after the CMP process.
  • the substrate having a planarized copper-containing surface forms a part of an article of manufacturing.
  • the substrate is integrated in an article as a copper- containing element.
  • Exemplary articles of manufacturing which may incorporate copper-containing surfaces or copper-containing elements in their final form or in an intermediate prepared during manufacturing the articles include, but are not limited to, an integrated circuit, a wafer-based device, 3D integration and advanced interconnects.
  • wafer-based device referred to a section of a wafer on which a copper-containing electronic device has been produced by using the CMP process and composition presented herein among other processes and manufacturing steps.
  • an article which includes a wafer-based device having a copper- containing element, wherein the copper-containing element is being produced by the process presented herein; the article is characterized by a copper dishing value of less than 110 nanometer or less than 50 nm and a height range (Z range) value of less than 20 nm or less than 15 nm.
  • a "copper-containing element” describes an element in an article that comprises at least 50 weight percents copper, at least 60 weight percents, at least 70 weight percents, at least 80 weight percents, at least 90 weight percents copper, or is made of 100 percents copper.
  • the element is the final product is obtained upon planarizing an intermediate that had a copper-containing surface.
  • the final product, the article has a copper-containing surface.
  • the CMP slurry composition embodies and incorporates a balanced combination of two opposing chemical processes, copper removal by corrosion and aberration, versus rapid copper passivation to corrosion. This non-trivial balance is achieved by specific concentrations of agents which accomplish the chemical processes, namely oxidizers such as H 2 0 2 and complexing agents such as glycine, opposing sorbate as a passivating/inhibiting/re- passivating agent.
  • oxidizers such as H 2 0 2
  • complexing agents such as glycine
  • the effect of this non-trivial balance can be measured by the rate of copper removal, indicating an effective corrosion process; a low dishing rate, indicative of the rapid re-passivation of the passivation process; low Z-range of the surface, indicating smoothness; and low resistance of test copper sheets, indicating an overall successful CMP process, which ultimately predicts and ensures a low rate of disqualified end-product devices on a wafer.
  • compositions, method or structure may include additional ingredients, steps and/or parts, but only if the additional ingredients, steps and/or parts do not materially alter the basic and novel characteristics of the claimed composition, method or structure.
  • a compound or “at least one compound” may include a plurality of compounds, including mixtures thereof.
  • range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the invention. Accordingly, the description of a range should be considered to have specifically disclosed all the possible subranges as well as individual numerical values within that range. For example, description of a range such as from 1 to 6 should be considered to have specifically disclosed subranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6 etc., as well as individual numbers within that range, for example, 1, 2, 3, 4, 5, and 6. This applies regardless of the breadth of the range.
  • the term "method" refers to manners, means, techniques and procedures for accomplishing a given task including, but not limited to, those manners, means, techniques and procedures either known to, or readily developed from known manners, means, techniques and procedures by practitioners of the chemical, pharmacological, biological, biochemical and medical arts. It is appreciated that certain features of the invention, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the invention, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable subcombination or as suitable in any other described embodiment of the invention. Certain features described in the context of various embodiments are not to be considered essential features of those embodiments, unless the embodiment is inoperative without those elements.
  • Cyclic polarization curves were obtained by reversing the scan direction once the recorded anodic potential reached a value of 1 V. Polishing Tool and Set-up:
  • CMP of patterned wafers was carried out on an Applied Materials Mirra Mesa polishing tool for more accurate polishing. CMP was performed with 25 % over- polishing time.
  • XPS X-ray photoelectron spectroscopy
  • Thermofisher a theta 300 Mkll system (Thermofisher), having Al -Ka X-radiation (1486.6 eV)).
  • the in-depth film composition and distribution of the oxidation state across the protective film on the copper surface were determined by an Angle Resolved XPS technique (ARXPS).
  • ARXPS Angle Resolved XPS
  • High resolution profilometer HRP
  • a KLA-Tencor high resolution profilometry tool was used to scan various structures and arrays on the patterned wafers in order to construct the relative surface topography and evaluate dishing values. Dishing values were measured on 80 ⁇ Van Der Pauw (VDP) structures.
  • R h0 is the resistivity of the sample. Sheet resistance values were normalized to a maximum value obtained in sample that was polished with low sorbate concentration.
  • Figure 1 is a comparative plot showing 2 anodic polarization curves of copper obtained by polarizing a copper electrode in a solution containing one gram per liter (g/L) of potassium sulfate (1 g/L K 2 S0 4 , curve denoted 'a'), and from polarizing copper in a solution containing one gram per liter of potassium sulfate and 8 grams per liter of potassium sorbate (curve denoted b").
  • Figure 2 is a comparative plot showing three anodic polarization curves of copper obtained by polarizing a copper electrode in a solution containing 4 g/L potassium sorbate (K-sorbate), each further containing either 1 g/L of potassium sulfate (K-sulfate), 1 g/L glycine, or 1 g/L potassium acetate (K-acetate).
  • K-sorbate potassium sorbate
  • Figure 3 is a comparative plot showing 4 anodic polarization curves of copper obtained by polarizing a copper electrode in a solution containing 1 g/L Na 2 S0 4 and 10 g/L glycine, and containing no hydrogen peroxide (curve denoted ⁇ 1 ), 0.3 percents by volume hydrogen peroxide (curve denoted '2"), 1 percents by volume hydrogen peroxide (curve denoted '3 1 ), and 5 percents by volume hydrogen peroxide (curve denoted '4').
  • the decrease in the I Cor r rnay be attributed to the adsorption of hydroxyl radicals at the copper surface.
  • the radicals are formed due to the decomposition of H 2 0 2 catalyzed by Cu-glycine complexes.
  • the increase in the radicals' concentration gradually expanded the passivity region from a value of 0 mV(scE) (in 0.3 percent by volume H 2 0 2 ) to 75 mV ( scE ) (in 1 percent by volume H 2 0 2 ) and 137 mV(scE ) (in 5 percent by volume H 2 0 2 ).
  • Figures 4A-F present high resolution SEM micrographs ( Figures 4A, 4C and 4E, left column) alongside with plots presenting potentiostatic measurements ( Figures 4B, 4D and 4F respectively, right column), obtained from a copper electrode polarized at 0.3 V(scE) for 200 seconds in a solution containing 1 g/L Na 2 S0 4 , and lOg/L glycine and containing 0.3 percents by volume hydrogen peroxide (Figures 4A-B), 1 percents by volume hydrogen peroxide ( Figures 4C-D), or 5 percents by volume hydrogen peroxide ( Figures 4E-F).
  • the surface morphology correlates with the current densities recorded from the potentiostatic measurements, wherein the solution containing 0.3 percent by volume H 2 0 2 , the copper surface was hardly covered with copper oxides/hydroxides. Pits (appearing as black spots in the SEM micrograph Figures 4A and 4C) were observed at the copper surface, indicating a local dissolution of the copper. In the solution containing 1 percent by volume H 2 0 2 , the copper surface was covered with copper oxides/hydroxides, which appeared as flakes structures. Interestingly, these structures did not cover the copper surface completely and pits caused by dissolved copper were seen at the surface.
  • the three regions of H 2 0 2 concentrations were selected to illustrate the effect of the sorbate at each region.
  • Figures 5A-F present comparative plots of potentiodynamic curves ( Figures 5A, 5C and 5E, left column) alongside with comparative plots of potentiostatic curves at 300 mV ( Figures 5B, 5D and 5F respectively, right column), obtained from a copper electrode polarized in solutions containing 1 g/L Na 2 S0 4 , 10 g/L glycine and containing 0.3 percents by volume hydrogen peroxide ( Figures 5A-B), 1 percents by volume hydrogen peroxide ( Figures 5C-D), or 5 percents by volume hydrogen peroxide ( Figures 5E-F), and further containing different K-sorbate concentrations as indicated in each plot.
  • Potentiostatic measurements performed at a potential value of 300 mV illustrate the stability of the formed protective film over time.
  • the anodic currents increased.
  • the addition of 4 g/L sorbate led to a decrease in the anodic currents.
  • the anodic currents increased, indicating on the low stability of the protective film, possibly because of a low coverage of the sorbate protective film at the copper surface.
  • solutions having sorbate concentrations of 16 and 64 g/L the anodic currents decreased and remained constant during the whole exposure time. This behavior indicates the formation of a highly protective layer at the copper surface.
  • CMP removal rate was studied in slurries containing various concentrations of sorbate (from none to 64 g/L), and peroxide (0.1 to 10 percent by volume), and a fixed amount of glycine (10 g/L).
  • Figure 6 presents a comparative plot showing the copper removal rate as a function of sorbate content in slurries containing 10 g L glycine with low or intermediate concentrations of H 2 0 2 .
  • Table 2 presents copper removal rate values as a function of sorbate content in slurries containing 10 g/L glycine with high H 2 0 2 concentration.
  • the presented data illustrate a typical trend in the copper removal rate as a function of the H 2 0 2 concentration in the slurry. This trend consists of three regions: (i) low concentrations of H 2 0 2 (0.1-0.3 percent by volume), (ii) intermediate concentrations of H 2 0 2 (0.3-1 percent by volume) ; and (iii) high concentrations of H 2 0 2 (5-10 percent by volume).
  • any addition of H 2 0 2 led to an increase in the copper removal rate until a maximum value was obtained at a concentration of 0.3 percent by volume H 2 0 2 .
  • further addition of H 2 0 2 decreased the copper removal rate up to a concentration of 1 percent by volume H 2 0 2 .
  • the increase in the sorbate concentration from 4 g/L to 64 g/L gradually decreased the removal rates values at both low and intermediate peroxide concentrations.
  • Figure 7 presents a comparative plot of three copper removal rate curves using CMP slurries containing three different H 2 0 2 concentrations and containing K-sorbate at a concentration of 16 g/L, each curve plotted as a function of glycine concentration and 3 different H 2 0 2 concentrations.
  • the copper removal rate exhibited a different behavior according to the H 2 0 2 concentration in the slurry.
  • increasing the glycine concentration from 1 to 20 g/L increased the cooper removal rate.
  • a further increase in the glycine concentration had an insignificant effect on the removal rate, even at high glycine concentrations.
  • a further increase in the glycine concentration gradually increased the removal rate.
  • Table 3 presents the copper removal rate values obtained in slurries containing different glycine and sorbate concentration ratios and 0.3 percent by volume H 2 0 2 .
  • the copper surface is composed of mainly copper metal with Cu(I) species as cuprous oxide. Therefore, oxidized copper ions are available at the exposed copper surface to react with either the glycine to form Cu-glycine soluble complexes, or with the sorbate to form a protective Cu-sorbate layer.
  • the copper surface is composed of both Cu(I) and Cu(II) species (partially covered with Cu(I) and Cu(II) oxides/hydroxides).
  • increase in the concentration of glycine above 20 g/L led to a gradual increase in the copper removal rate values. This behavior suggests that the copper removal rate is determined by relative reaction rates of Cu-glycine complexation, Cu-sorbate and copper oxides/hydroxides formation.
  • X-ray photoelectron spectroscopy (XPS) analysis of the copper surface was performed in order to determine the surface composition subsequent to the CMP process.
  • the samples were polished with CMP slurries containing 64 g/L K-sorbate, 10 g/L glycine and varying concentrations of H 2 0 2 .
  • Figures 8A-D present comparative plots of high-resolution X-ray photoelectron spectra (XPS) of Cu2p 3/2 obtained from copper surface subsequent to CMP process in slurries containing 64 g L K-sorbate, 10 g L glycine and three different H 2 0 2 concentrations (0.1, 1 and 10 percent by volume presented in Figures 8C, 8B and 8 A respectively) and recorded at a take-off angle of 12° relative to the copper surface, whereas etched copper sample spectrum serves as a reference ( Figure 8D).
  • XPS X-ray photoelectron spectra
  • the Cu2p 3/2 spectrum recorded in the etched sample is composed of a single peak located at 932.8eV and attributed to both a cuprous oxide and a metallic copper.
  • H 2 0 2 concentration increased from 0 to 0.1 and to 1 percent by volume, an energy shift of 0.5 eV in the Cu2p 3/2 main peak was observed.
  • a further increase in the H 2 0 2 concentration to 10 percent by volume shifted the main peak back to 932.8 eV, as recorded in the etched sample.
  • the presence of the satellite peaks at 940-945 eV verify the presence of Cu(II) species.
  • the additional peaks are positioned at binding energies of 934.8 and 936.3 eV. These peaks can be referred to CuO and Cu(OH) 2 , respectively. Attributing these peaks to CuO and Cu(OH) 2 can be further verified by Ols spectra.
  • the additional peaks located at 933.8eV and 935. leV can be unambiguously attributed to CuO and Cu(OH) 2 , respectively. Further investigations of the Ols spectra provided additional information on the oxidized copper species, which coincided with the results obtained from the Cu2p 3/2 spectra.
  • Figures 9A-C present comparative plots of high-resolution X-ray photoelectron spectra (XPS) of Ols obtained from copper surface subsequent to CMP process in slurries containing 64 g/L K-sorbate, 10 g/L glycine and different H 2 0 2 concentrations (0.1, 1 and 10 percent by volume presented in Figures 9C, 9B and 9 A respectively).
  • XPS X-ray photoelectron spectra
  • the peak related to Cu 2 0 shifted to lower binding energies, related to CuO.
  • the precise intensity of the cupric oxide layer (CuO) cannot be extracted from the Ols spectrum due to a small chemical shift, compared with Cu 2 0.
  • broadening of the peak to lower binding energies associated with CuO may indicate the formation of CuO, as can be clearly seen in the Cu2p 3 2 spectra.
  • Figures 10A-B present Cu2p 3 2 ( Figure 10A) and Ols (Figure 10B) X-ray photoelectron spectra (XPS) at different take-off angles taken from the surface of a copper sample subsequent to CMP process in slurry containing 64 g/L K-sorbate, 10 g/L glycine, and low H 2 0 2 concentration of 0.1 percent by volume.
  • the take-off angle of the bulk mode (69° related to the copper surface) provides maximum information from the bulk.
  • the Cu2p 3/2 spectrum recorded at a take-off angle of 69° consists of only the main peak shifted to a higher binding energy with two correlating peaks assigned to Cu 2 0 and C-0 in the Ols spectrum.
  • the energy shift in the Cu2p 3 2 spectrum was attributed to an induced dipole at the copper surface.
  • the dipole may have resulted due to the formation of a copper-oxygen association between the polished copper surface and an oxygen atom in the sorbate molecule. This association may be formed adjacent to the exposed surface areas during the CMP in 0.1 and 1 percent by volume H 2 0 2 slurries.
  • Figures 11A-B present Cu2p 3/2 ( Figure 11A) and Ols (Figure 11B) X-ray photoelectron spectra (XPS) at different take-off angles taken from the surface of a copper sample subsequent to CMP process in slurry containing 64 g/L K-sorbate, 10 g/L glycine, and high H 2 0 2 concentration of 10 percent by volume.
  • XPS X-ray photoelectron spectra
  • the peak fit of the Cu2p 3/2 spectra indicates that the relative intensity of Cu(OH) 2 and CuO increased with a decrease in the take-off angle, indicating a higher concentration of Cu(OH) 2 and CuO at the top surface level of the protective layer.
  • a decrease in the relative intensities of Cu 2 0 peaks was observed as the take-off angle decreased, indicating a higher in-depth concentration of Cu 2 0.
  • the results obtained from the Ols spectra coincide with those obtained in the Cu2p 3 2 spectra.
  • Figure 12 presents comparative plots of extrapolated data estimated from the intensity ratio (Io/I m ) between the oxide and the metal peaks obtained in XPS measurements, showing the oxide thickness developed on copper as a function of H 2 0 2 content in the slurry.
  • the oxide thickness was estimated using Strohmerier [Strohmeier, B.R., Surf. Int. Anal. 15 (1990), 51] and Seah [Seah, M. et al, Surf. Int. Anal. 1 (1979), 2] equations.
  • the copper oxide layer is mainly composed of Cu 2 0 with a thickness of about 0.8 nm at low concentration of H 2 0 2 (0.1 percent by volume), while at a high H 2 0 2 concentration (10 percent by volume), the copper oxide layer is mainly composed of CuO with a thickness of about 2.8 nm.
  • the oxide composition is modified with an increase in the H 2 0 2 concentration, but also its thickness. The thickness and layer composition would have a major effect on both the copper dissolution and passivation during the CMP process.
  • Static contact angles of de-ionized water at the planarized copper surface were used to indicate and quantify the change in the surface tension of a copper-containing substrate.
  • Figures 13A-D present contact angle measurements of a water droplet at a copper wafer surface ( Figures 13A and 13B) and SEM images ( Figures 13C and 13D) obtained from the copper wafer surface, subsequent to CMP process in slurries containing 64 g L K-sorbate and 10 g/L glycine with H 2 0 2 concentrations of 0.1 percent by volume ( Figures 13 A and 13C) and of 5 percents by volume ( Figures 13B and 13D).
  • the increase in the H 2 0 2 concentration from 0.1 to 5 percent by volume decreased the contact angle from 68 ° to 53 °, indicating that there was a change in the surface chemistry, which altered its hydrophilicity.
  • Figure 14 presents comparative plots showing the XPS intensities ratio of
  • Cu(OH) 2 and CuO normalized to the Cu° peak in the Cu2p 3/2 spectrum, as measured at an angle of 12° related to the copper surface subsequent to polishing with slurries containing 64 g/L K-sorbate, 10 g/L glycine and different H 2 0 2 concentrations.
  • Figures 15A-D present contact angle measurements of a water droplet at a copper wafer surface ( Figures 15A and 15B) and SEM images ( Figures 15C and 15D) obtained from the copper wafer surface, subsequent to CMP process in slurries containing 10 g/L glycine and 0.1 percent by volume H 2 0 2 with sorbate concentrations of 4 g/L ( Figures 15 A and 15C) and of 64 g/L ( Figures 15B and 15D).
  • the increase in the sorbate concentration from 4 to 64 g/L in slurries containing low concentration of H 2 0 2 increased the contact angle from 80° to 99°, indicating the presence of the fatty acid (sorbate) at the copper surface.
  • the sorbate molecule is a 6-carbons fatty acid chain having hydrophobic alkyl tails; thus, the decrease in the wettability, as sorbate concentration increased was most probably due to the sorbate hydrophobic tail.
  • Figures 16A-D summarize the postulated processes occurring on the surface of copper when using the H 2 02-glycine slurries containing sorbate, as concluded from the experimental results discussed hereinabove.
  • Figures 16A-D are schematic diagrams depicting a theoretical surface of a copper sample polished by slurries having two concentrations of H 2 0 2 , 0.1-0.3 percent by volume (low H 2 0 2 , Figures 16A and 16B) and 5-10 percent by volume (high H 2 0 2 , Figures 16C and 16D), and two sorbate concentrations, 4g/L (low sorbate, Figures 16A and 16C) and 64 g/L (high sorbate, Figures 16B and 16D).
  • the sample depicted in Figure 16 A can be characterized by a copper removal rate of 240 nm/min and a contact angle of 80 °; the sample depicted in Figure 16B can be characterized by a copper removal rate of 83 nm/min and a contact angle of 99 °; the sample depicted in Figure 16C can be characterized by a copper removal rate of 234 nm/min and a contact angle of 51 °; and the sample depicted in Figure 16D can be characterized by a copper removal rate of 246 nm/min and a contact angle of 69 °.
  • the surface composition and thickness are being altered according to the constituent's concentration in the slurry, consequently determining copper removal rate and wafer performance.
  • Figure 17 presents a 4 x 5 array of 20 optical microscope images of copper surface subsequent to CMP process with CMP slurries containing different concentration of sorbate and H 2 0 2 and a fixed glycine concentration of 10 g/L, denoting the removal rate (RR in nanometer per minute, nm/min) values, wherein the concentration of H 2 0 2 increases column-wise 0.1, 0.3, 1, 5 and 10 percents by volume, and the concentration of sorbate increases row- wise 0, 4, 16 and 64 g/L.
  • the surface is mainly covered with islands, or possibly a thin layer, of cuprous oxides (0.8 nm) alongside Cu- sorbate species formed on top of the copper surface.
  • cuprous oxides 0.8 nm
  • the oxide layer does not cover the entire surface or is not thick enough to protect the copper surface.
  • the surface is characterized with a coarse roughness, indicating uncontrolled etching of copper.
  • the surface exhibits a finer topography with a reduced roughness.
  • the copper surface is mainly composed of cupric oxides and hydroxides (about 4 nm).
  • the roughness decreases as the etching of the surface occurs in a more controlled manner.
  • the oxide/hydroxide layer is additionally covered with adsorbed sorbate anions and the surface exhibits an improved roughness, even with the addition of low sorbate concentrations.
  • AFM measurements were performed on copper surface subsequent to CMP with slurries containing 10 g L glycine, 5 percent by volume H 2 0 2 and different concentrations of sorbate, in order to quantify the effect of the sorbate concentration on the roughness during the copper CMP process itself.
  • Figures 18A-B present AFM micrographs of two copper samples subsequent to polishing with slurries having 4 g/L ( Figure 18A) and 64 g/L ( Figure 18B), wherein the roughness of the sample treated with 4 g/L sorbate exhibited a Z range of 37.319 nm, an
  • Patterned wafer are used extensively by semiconductor capital equipment manufacturers to develop new wafer fabrication processes and maintain existing ones.
  • a typical CMP characterization wafers are used to evaluate the performance of various CMP products such as slurries.
  • Patterned test wafers differ from blanket film test wafers because they include additional process steps to create patterns on the blanket film and contain several arrays with copper/barrier features of different sizes and spacing.
  • the die pattern can include many test structures, including lines, spaces, holes and different features. Hence, a patterned test wafer simulates the behavior of device wafers processed under similar conditions.
  • Wafer performance and polishing uniformity are Wafer performance and polishing uniformity:
  • Figures 19A-D present optical microscope images of copper bond pads (100 x 100 ⁇ ) subsequent to copper and barrier CMP, using copper CMP slurries having 4 g/L sorbate ( Figures 19A and 19C) or 64 g L sorbate ( Figures 19B and 19D), polished by an edge die ( Figures 19A and 19B) or a center die ( Figures 19C and 19D), demonstrating the surface finish and polishing uniformity as a function of the sorbate concentration.
  • the film produced at the copper surface is composed of hydroxides, oxides and sorbate multilayer in an overall thickness of 4-10 nm. As no monolayer is being formed, one would need a sufficient dissolved sorbate in the solution in order to establish a full coverage (and hence an efficient protection) of the copper surface.
  • Figures 20A-D present SEM images of 90 nm copper wiring, showing top view
  • Figures 20A and 20B and side view ( Figures 20C and 20D), polished with CMP slurries having 4 g L sorbate ( Figures 20A and 20C) or 64 g/L sorbate ( Figures 20B and 20D), wherein copper corrosion sites are marked by arrows in Figure 20C.
  • FIG. 21 presents the dishing values of copper van der-Pauw (VDP) structures
  • dishing values decrease substantially with increase in sorbate concentration.
  • the decrease in dishing values with increased sorbate concentration indicates that sorbate anions were adsorbed at the oxide layer and covered a larger portion of the surface thus provided an enhanced protection to the copper lines and structures.
  • Figure 22 presents a box plot of normalized copper sheet resistance values at different copper densities (PD) as a function of sorbate concentrations, wherein the sheet resistance values are normalized to a maximum value obtained in sample that was polished with a slurry having a sorbate concentration of 4 g/L.
  • PD copper densities
  • Figure 23 presents a box plot of normalized copper sheet resistances values as a function of sorbate concentration in the slurry measured from different line widths, wherein the sheet resistance values are normalized to a maximum value obtained from sample polished with a slurry having a sorbate concentration of 4 g/L.
  • Figure 24 summarizes the CMP process with respect to sorbate concentration in a 5 percent by volume H 2 0 2 -glycine based slurry as concluded from the experimental results presented hereinabove.
  • Figures 24A-D are schematic diagrams depicting a theoretical surface of a copper sample polished by slurries having two sorbate concentrations, 4 g/L (low sorbate, Figures 24A and 24C) and 64 g/L (high sorbate, Figures 24B and 24D), in two polishing stages, the main CMP process ( Figures 24A and 24B) and the over-polishing stage ( Figures 24C and 24D).
  • the sample depicted in Figure 24A can be characterized by a height range of 37 nm; the sample depicted in Figure 24B can be characterized by a height range of 13 nm; the sample depicted in Figure 24C can be characterized by a dishing value subsequent to copper CMP process of 140 nm; and the sample depicted in Figure 24D can be characterized by a dishing value subsequent to copper CMP process of 50 nm.
  • sorbate anions are adsorbed to the copper surface at the recessed areas, assisting in achieving a better planarization, while at low sorbate concentration the recessed areas are not well protected and uncontrolled etching occurs along the copper lines.
  • the recessed areas are protected and etching is restrained, therefore, at high sorbate concentrations, lower height range and RMS roughness values are recorded (AFM measurements).
  • the copper structures should be protected to avoid further etching of the planarized structures, which could result in open circuit of the device.
  • the protective layer exposed to the slurry is composed of mainly CuO and Cu(OH) 2 .
  • copper oxides and hydroxides are not well protected and dissolve in the presence of glycine and acidic media, which is typically used in subsequent steps, further down in the production line of ICs.
  • copper oxides/hydroxides films are well covered with protective adsorbed sorbate anions therefore, lower dishing values are obtained.
  • FIGS 25A-D Electron micrographs were taken before and after the post-treatment with sorbate, and the results are presented in Figures 25A-D.
  • Figures 25A-D are electron micrographs at a magnification of 8,000X (Figure 25 A), 64,000X (Figure 25B), 10,000X (Figure 25C) and 50,000X ( Figure 25D) of copper-containing wafer samples subjected to CMP before post- treatment with 0.4 M sorbate ( Figures 25A and 25B) and after post-treatment with 0.4 M sorbate ( Figures 25C and 25D), showing significant decrease in the silica particles adhesiveness as a result of the surface passivation effect imparted to the copper surface by sorbate.
  • Results obtained from electrochemical, contact angle, XPS and CMP studies presented hereinabove provide an insight into the mechanism involved in copper CMP conducted with H 2 0 2 -glycine slurry containing sorbate as an inhibitor, and allow determination of a CMP slurry composition with superior performance as compared to currently known CMP slurries.
  • Electrochemical studies show that the addition of sorbate to H 2 0 2 -glycine solutions expands the passivity region by providing an enhanced passivity due to the formation of both an oxide and a sorbate protective film at the copper surface.
  • XPS results reveal that both the surface composition and the thickness of the protective layer alter as H 2 0 2 concentration increases.
  • an exemplary beneficial sorbate anion concentration in the CMP slurry composition is 0.4 M
  • the concentration of glycine is 20 grams/L
  • a concentration of hydrogen peroxide is 10 percents by volume.

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Organic Chemistry (AREA)
  • Physics & Mathematics (AREA)
  • Condensed Matter Physics & Semiconductors (AREA)
  • General Physics & Mathematics (AREA)
  • Manufacturing & Machinery (AREA)
  • Computer Hardware Design (AREA)
  • Microelectronics & Electronic Packaging (AREA)
  • Power Engineering (AREA)
  • Mechanical Treatment Of Semiconductor (AREA)

Abstract

Disclosed are copper CMP compositions containing a high concentration of sorbate anion. Further disclosed are processes of planarizing a copper-containing surface using the compositions, as well as substrates and articles having planarized copper-containing elements produced by the disclosed processes.

Description

S ORB ATE-CONTAINING COMPOSITIONS FOR USE IN COPPER CHEMICAL
MECHANICAL PLANARIZATION .
FIELD AND BACKGROUND OF THE INVENTION
The present invention, in some embodiments thereof, relates to chemical- mechanical planarization (CMP) and, more particularly, but not exclusively, to compositions that are useful for chemical-mechanical planarization (CMP) of copper- containing surfaces, to processes utilizing same and to products produced by these processes.
Wafer-based electronic devices are manufactured with several levels of interconnected levels. As the number of interconnect levels increases, planarization of the previous layers is required to ensure a flat surface prior to subsequent lithography. Without planarization, the levels would become increasingly crooked and extend outside the depth of focus of available lithography, interfering with the ability to pattern. Chemical Mechanical Planarization (CMP) is the primary processing method to achieve such planarization.
CMP is a process whereby chemical reactions increase the mechanical removal rate of a material, typically a metal. CMP is commonly used to polish off high spots on wafers or other features deposited on wafers, flattening the feature or the wafer, referred to as planarization. CMP is executed by use of slurries containing abrasive particles that effect the mechanical polishing, and reagents that effect the chemical reactions. The chemical reaction that increases the mechanical removal rate is commonly tailored to provide a higher removal rate of one material versus another material. The chemical action in CMP helps to achieve a higher removal selectivity of one material to another than a strictly mechanical process would provide. Another requirement of the chemical reactions and reagents is that it would leave the polished layer passive to further uncontrolled corrosion (re-passivation); hence, the corrosive reactivity of the CMP reagents should be balanced with rapid re-passivation of the polished surface. Each component in a balanced CMP slurry serves one of the main CMP objectives and affects, and to some extent negates the others, as is further discussed in detail hereinbelow. Thus, some of the requirements of a well balanced CMP process include a relatively high removal rate (e.g. >200 nm/minute), for affording a smooth polished surface, namely high planarization efficiency; instantaneous self-passivation of the freshly polished surface (which can optionally be achieved by a protective film with high protective characteristics); and rapid re-passivation of surface sections, where the passivation film is removed by the mechanical action.
Copper is widely used in the manufacturing of integrated circuits (IC) owing to its low electrical resistance (1.7 μΩ/cm) and excellent electromigration resistance. The integration of copper into IC is performed by a damascene technique, which involves CMP for removing the excess copper subsequent to wafer plating, in order to form a planar surface for subsequent fabrication of additional levels of chip wiring. A high performance of the CMP process allows a successful fabrication of multilevel layers.
Typical copper CMP slurries consist of abrasives, such as silica particles, dispersed in an aqueous solution containing an oxidizer, such as hydrogen peroxide.
Hydrogen peroxide (H202) is a powerful oxidizer and can be effectively used in various solutions in a wide range of pH values.
CMP slurries typically also require a complexing agent to dissolve the oxidized species and a corrosion inhibitor to form a passivating layer to protect the lower regions from excessive etching. Glycine is an example of a widely used complexing agent, typically employed in acidic slurries. In peroxide-containing slurries, apart from glycine, few other complexing agents have also been proposed. These include, for example, acetic acid, ethylene diamine, glycolic acid, phthalic acid salt, citric acid, and oxalic acid.
Copper CMP can be performed in acidic or neutral or basic medium. In silicon wafer and IC manufacturing, a high selectivity of Cu/Si02 polish rate is desirable to minimize erosion losses, as can be achieved with acidic CMP slurry. However, acidic slurries are disadvantageous due to possible corrosion of the polishing equipment by the slurry. In the alkaline region, ammonia and ammonium salt based slurries have been investigated but the reported Cu/Si02 polish rate selectivity was found to be an order of magnitude lower than that achievable in acidic slurries.
Glycine and H202 containing slurries, which are typically utilized in acidic medium, were also investigated in alkaline regime where the pH was maintained by buffers; however, the polish rate of silicon dioxide with that slurry has not been studied extensively.
U.S. Patent Application No. 20070163677 teaches copper CMP slurry compositions which has a pH that ranges from 9 to 13, and which are devoid of ammonium-containing complexing agents.
During copper CMP process, a short period of over polishing is needed in order to avoid electrical shorts between the copper lines across the wafer. However, dishing of copper features and erosion of the dielectric materials occur during such over- polishing step. Dishing reduces the thickness of wide copper features, leading to an increase in resistance and current density along the line. This in turn, leads to reduced signal propagation along the line and to a higher risk for electromigration phenomenon. In addition, it can complicate further lithography steps and the integration of additional metal layers.
If the embedded copper structures and lines are not well protected, the following steps of the IC manufacturing process would ultimately worsen corrosion phenomena which appear during the copper CMP, eventually leading to a device failure. Furthermore, dishing has been categorized as a parameter in CMP process that must be controlled. The dishing phenomenon was attributed to unbalanced copper passivation, uncontrolled dissolution of the copper structures and to the bending of the used CMP pad.
CMP slurry performance and its influence on dishing thus depend, at least in part, on the use of a corrosion inhibitor and on its characteristics. The inhibitor is required to perform during three main stages; the copper CMP process itself, the over- polishing stage and the next stage down the line of the IC manufacturing process, which may employ different chemical/electrochemical media and different mechanical forces. The inhibitor's ability to protect copper surface in such a dynamic process is determined by a few factors; the inhibitor's passivation/re-passivation kinetics, and the protective film stability in different pH regions and in the presence of oxidizing and complexing agents which are used throughout the IC manufacturing process.
Copper-corrosion inhibitors (also referred to herein and in the art as copper- etching inhibitors) have been described in the art. However, the integration of copper- corrosion inhibitors that perform well in solutions or other static media, in a dynamic process such as CMP is a non-trivial challenge. A copper-corrosion inhibitor is employed in a CMP slurry in order to obtain a planarized and passivated surface by forming a protective film both at the recessed areas during the CMP process, and at the whole planarized surface once the CMP is completed (post-CMP). A successful performance of a copper-etching inhibitor in a static medium therefore cannot predict a successful performance of the inhibitor is a dynamic process such as CMP.
The currently most widely used copper-etching inhibitor in CMP processes is benzotriazole (BTA). Benzotriazole (BTA) is a very effective inhibitor of copper corrosion in acidic, neutral and basic media; however, it is reported to be slow in its inhibiting effects and toxic, besides posing other post-CMP challenges.
Although the inhibitor characteristics have a decisive impact on the CMP process, very few CMP studies reported the use of copper-corrosion inhibitors other than BTA. Other additives proposed as copper corrosion inhibitors include, for example, citric acid, hydrazine, amino tetrazole, potassium sorbate [Abelev, E. et al, Electrochim. Acta 52 (2007), p. 1975], and imidazole. Other agents used for passivating copper are taught in, for example, U.S. Patent Nos. 5,770,095 and 6,569,350 and U.S. Patent Application No. 20070163677.
Potassium sorbate (potassium (2E,4E)-hexa-2,4-dienoate, CAS number 24634- 61-5, 150.22 grams per mol, hereinafter K-sorbate) is the potassium salt of sorbic acid, a naturally occurring carboxylic acid, which finds its primary use in the food and personal care industries as the preservative Ε202. Sorbate salts are generally accepted as safe and non-toxic, and are typically very soluble in water, up to 58.2 % at 20 °C in the case of potassium sorbate. Being available at relatively low cost and being safe, sorbate salts such as potassium sorbate, are highly desirable alternative reagents in many industries.
Prior art publications [Y. Ein-Eli, et al., Electrochem. Solid-State Lett. 9 (2006), B5; and E. Abelev, et al., Langmuir 23 (2007), p. 11281] investigated potassium sorbate as an inhibitor of copper corrosion in aqueous medium. Other publications discuss CMP slurries containing potassium sorbate [Ein-Eli, Y. et al., Electrochim. Acta 52 (2007), p 1825; Abelev, E. et al., Electrochim. Acta 52 (2007), p 1975; Abelev, E. et al., Electrochim. Acta 52 (2007), p 5150; Ein-Eli, Y. et al., Electrochim. Acta 49 (2004), p. 1499; and Ein-Eli, Y. et al., J. Electrochem. Soc. 151 (2004), G236]. Chemical mechanical planarization of copper disks in hydrogen peroxide and L- arginine based alkaline slurry was reported by Y.N. Prasad and S. Ramanathan, [Electrochimica Acta, 52 (2007), pp. 6353-6358]. This study screened common inhibitors by static etch rate experiments, and found only BTA and uric acid to be effective in the alkaline pH range, while potassium sorbate was found moderately effective when used in a low concentration, and almost ineffective when used in high concentrations.
SUMMARY OF THE INVENTION
The present invention, in some embodiments thereof, relates to highly effective chemical-mechanical planarization (CMP) of copper using sorbate as a copper passivation agent in high concentration of more than 0.2 molar. These CMP compositions surpass known compositions in many criteria in terms of process and end- product characteristics, such as copper removal rate, dishing, smoothness, and post- CMP conditioning of the planaraized copper surface.
Hence, according to an aspect of embodiments of the invention presented herein, there is provided a copper CMP composition that includes hydrogen peroxide, glycine and a sorbate anion, wherein the concentration of sorbate in the composition is at least 0.2 mol per liter.
According to another aspect of embodiments of the invention presented herein, there is provided a copper CMP composition that includes:
a peroxide in a concentration that ranges from 0.1 percent by volume to 10 percent by volume;
a complexing agent in a concentration that ranges from 1 gram per liter to 40 grams per liter; and
a sorbate anion in a concentration of at least 0.2 mol per liter.
In some embodiments, the concentration of the peroxide ranges from 5 percent by volume to 10 percents by volume.
In some embodiments, the peroxide is hydrogen peroxide.
In some embodiments, the concentration of the complexing agent ranges from 5 grams per liter to 30 grams per liter. In some embodiments, the complexing agent is selected from the group consisting of an amino acid, a monoamine-containing compound and a diamine- containing compound.
In some embodiments, the complexing agent is glycine.
In some embodiments, the concentration of sorbate is at least 1.7 mol per liter.
In some embodiments, the source of the sorbate anion is potassium sorbate.
In some embodiments, the concentration of hydrogen peroxide ranges from 0.1 percent by volume to 10 percents by volume, or from 5 percent by volume to 10 percents by volume, or from 0.3 percent by volume to 1 percent by volume.
In some embodiments, the concentration of glycine ranges from 5 grams per liter to 40 grams per liter, or from 5 grams per liter to 30 grams per liter, or from 10 grams per liter to 30 grams per liter.
In some embodiments, the concentration of the sorbate anion ranges from 0.4 mol per liter to 9 mol per liter, a concentration of glycine is 20 grams per liter, and a concentration of hydrogen peroxide ranges from 5 percent by volume to 10 percents by volume.
In some embodiments, the composition presented herein has a pH that ranges from 7 to 10, or a pH of 8.5.
In some embodiments, the composition presented herein further includes abrasive particles.
In some embodiments, the abrasive particles are selected from the group consisting of silica, alumina, titania, zirconia, ceria, germanium oxide and mixtures thereof.
According to another aspect of embodiments of the invention presented herein, there is provided a process of chemical-mechanical planarization of a copper-containing surface, which is effected by contacting a substrate having a copper-containing surface with the copper CMP slurry composition presented herein, thereby planarizing the copper-containing surface.
In some embodiments, the process includes, subsequent to the contacting, applying a polishing pad onto the substrate.
In some embodiments, the process is characterized by a copper removal rate of at least 200 nanometer per minute. In some embodiments, the process is characterized by a planarized copper- containing surface that has a copper dishing value of less than 110 nanometer.
According to another aspect of embodiments of the invention presented herein, there is provided a substrate having a planarized copper-containing surface obtained by the process presented herein.
In some embodiments, the planarized copper-containing surface is characterized by a height range (Z range) value of less than 20 nm.
In some embodiments, the planarized copper-containing surface is characterized by a water-to-surface contact angle of at least 80 °.
In some embodiments, the substrate forma a part of an article that comprises a copper-containing element.
In some embodiments, the article is selected from the group consisting of an integrated circuit, a device wafer, a 3D integration device and an advanced interconnect device.
According to another aspect of embodiments of the invention presented herein, there is provided an article that includes a wafer-based device having at least one copper-containing element, the copper-containing element being produced by the process presented herein.
In some embodiments, the article is characterized by a copper dishing value of less than 110 nanometers and/or a height range (Z range) value of less than 20 nm.
Unless otherwise defined, all technical and/or scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the invention pertains. Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of embodiments of the invention, exemplary methods and/or materials are described below. In case of conflict, the patent specification, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and are not intended to be necessarily limiting. BRIEF DESCRIPTION OF THE DRAWINGS
The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee. Some embodiments of the invention are herein described, by way of example only, with reference to the accompanying drawings and images. With specific reference now to the drawings and images in detail, it is stressed that the particulars shown are by way of example and for purposes of illustrative discussion of embodiments of the invention. In this regard, the description taken with the drawings and images makes apparent to those skilled in the art how embodiments of the invention may be practiced.
In the drawings:
FIG. 1 is a comparative plot showing two anodic polarization curves of copper obtained by polarizing a copper electrode in a solution containing one gram per liter (g/L) of potassium sulfate (1 g/L K2SO4, curve denoted 'a1), and from polarizing copper in a solution containing one gram per liter of potassium sulfate and 8 grams per liter of potassium sorbate (curve denoted b1);
FIG. 2 is a comparative plot showing three anodic polarization curves of copper obtained by polarizing a copper electrode in a solution containing 4 g/L potassium sorbate (K-sorbate), each further containing either 1 g/L of potassium sulfate (K- sulfate), 1 g/L glycine, or 1 g/L potassium acetate (K-acetate);
FIG. 3 is a comparative plot showing four anodic polarization curves of copper obtained by polarizing a copper electrode in a solution containing 1 g/L Na2S04 and 10 g/L glycine, and containing no hydrogen peroxide (curve denoted Ί1), 0.3 percents by volume hydrogen peroxide (curve denoted '2"), 1 percents by volume hydrogen peroxide (curve denoted '3 and 5 percents by volume hydrogen peroxide (curve denoted '4');
FIGs. 4A-F present high resolution SEM micrographs (FIGs. 4A, 4C and 4E, left column) alongside with plots presenting potentiostatic measurements (FIGs. 4B, 4D and 4F respectively, right column), obtained from a copper electrode polarized at 0.3 (scE) for 200 seconds in a solution containing 1 g/L Na2S04 andlOg/L glycine and containing 0.3 percents by volume hydrogen peroxide (FIGs. 4A-B), 1 percents by volume hydrogen peroxide (FIGs. 4C-D), or 5 percents by volume hydrogen peroxide (FIGs. 4E-F); FIGs. 5A-F present comparative plots of potentiodynamic curves (FIGs. 5A, 5C and 5E, left column) alongside with comparative plots of potentiostatic curves at 300 mV (FIGs. 5B, 5D and 5F respectively, right column), obtained from a copper electrode polarized in solutions containing 1 g/L Na2S04, 10 g/L glycine and containing 0.3 percents by volume hydrogen peroxide (FIGs. 5A-B), 1 percents by volume hydrogen peroxide (FIGs. 5C-D), or 5 percents by volume hydrogen peroxide (FIGs. 5E-F), and further containing different K-sorbate concentrations as indicated in each plot;
FIG. 6 presents a comparative plot showing the copper removal rate as a function of sorbate content in slurries containing 10 g/L glycine with low or intermediate concentrations of H202;
FIG. 7 presents a comparative plot of three copper removal rate curves using CMP slurries containing three different H202 concentrations and containing K-sorbate at a concentration of 16 g L, each curve plotted as a function of glycine concentration and 3 different H202 concentrations;
FIGs. 8A-D presents comparative plots of high-resolution X-ray photoelectron spectra (XPS) of Cu2p3/2 obtained from copper surface subsequent to CMP process in slurries containing 64 g/L K-sorbate, 10 g/L glycine and three different H202 concentrations (0.1, 1 and 10 percent by volume presented in FIG. 8C, FIG. 8B and FIG. 8A respectively) and recorded at a take-off angle of 12° relative to the copper surface, whereas etched copper sample spectrum serves as a reference (FIG. 8D);
FIGs. 9A-C presents comparative plots of high-resolution X-ray photoelectron spectra (XPS) of Ols obtained from copper surface subsequent to CMP process in slurries containing 64 g/L K-sorbate, 10 g/L glycine and different H202 concentrations (0.1, 1 and 10 percent by volume presented in FIG. 9C, FIG. 9B and FIG. 9A respectively);
FIGs. 10A-B present Cu2p3 2 (FIG. 10A) and Ols (FIG. 10B) X-ray photoelectron spectra (XPS) at different take-off angles taken from the surface of a copper sample subsequent to CMP process in slurry containing 64 g/L K-sorbate, 10 g/L glycine and low H202 concentration of 0.1 percent by volume. The take-off angle of the bulk mode (69° related to the copper surface) provides maximum information from the bulk; FIGs. 11A-B present Cu2p3/2 (FIG. 11A) and Ols (FIG. 11B) X-ray photoelectron spectra (XPS) at different take-off angles taken from the surface of a copper sample subsequent to CMP process in slurry containing 64 g/L K-sorbate, 10 g/L glycine and high H202 concentration of 10 percent by volume;
FIG. 12 presents comparative plots of extrapolated data estimated from the intensity ratio (Io/Im) between the oxide and the metal peaks obtained in XPS measurements, showing the oxide thickness developed on copper as a function of H202 content in the slurry having 64 g/L K-sorbate and 10 g/L glycine;
FIGs. 13A-D present contact angle measurements of a water droplet at a copper wafer surface (FIGs. 13A and 13B) and SEM images (FIGs. 13C and 13D) obtained from the copper wafer surface, subsequent to CMP process in slurries containing 64 g/L K-sorbate and 10 g/L glycine with H202 concentrations of 0.1 percent by volume (FIGs. 13 A and 13C) and of 5 percents by volume (FIGs. 13B and 13D);
FIG. 14 presents comparative plots showing the XPS intensities ratio of Cu(OH)2 and CuO, normalized to the Cu° peak in the Cu2p3/2 spectrum, as measured at an angle of 12 ° related to the copper surface subsequent to polishing with slurries containing 64 g/L K-sorbate, 10 g/L glycine and different H202 concentrations;
FIGs. 15A-D present contact angle measurements of a water droplet at a copper wafer surface (FIGs. 15A and 15B) and SEM images (FIGs. 15C and 15D) obtained from the copper wafer surface, subsequent to CMP process in slurries containing 10 g/L glycine and 0.1 percent by volume ¾02 with sorbate concentrations of 4 g/L (FIGs. 15 A and 15C) and of 64 g/L (FIGs. 15B and 15D);
FIGs. 16A-D summarizes the postulated processes occurring on the surface of copper when using the H202-glycine slurries containing sorbate, as concluded from the experimental results discussed hereinabove. FIGs. 16A-D are schematic diagrams depicting a theoretical surface of a copper sample polished by slurries having two concentrations of H202, 0.1-0.3 percent by volume (low H202, FIGs. 16 A and 16B) and 5-10 percent by volume (high H202, FIGs. 16C and 16D), and two sorbate concentrations, 4g/L (low sorbate, FIGs. 16A and 16C) and 64 g/L (high sorbate, FIGs. 16B and 16D);
FIG. 17 presents a 4X5 array of 20 optical microscope images of copper surface subsequent to CMP process with CMP slurries containing different concentration of sorbate and H202 and a fixed glycine concentration of 10 g/L, denoting the removal rate (RR in nanometer per minute, nm/min) values, wherein the concentration of H202 increases column-wise 0.1, 0.3, 1, 5 and 10 percents by volume, and the concentration of sorbate increases row-wise 0, 4, 16 and 64 g/L;
FIGs. 18A-B present AFM micrographs of two copper samples subsequent to polishing with slurries having 4 g L (FIG. 18A) and 64 g/L (FIG. 18B) sorbate, wherein the roughness of the sample treated with 4 g/L sorbate exhibited a Z range of 37.319 nm, an RMS (Rq) of 1.621 nm, and a mean roughness (Ra) of 1.306 nm, and the sample treated with 64 g/L sorbate exhibited a Z range of 13.533 nm, an RMS (Rq) of 1.284 nm, and a mean roughness (Ra) of 0.99 nm;
FIGs. 19A-D present optical microscope images of copper bond pads (100 x 100 μπι2) subsequent to copper and barrier CMP, using copper CMP slurries having 4 g/L sorbate (FIGs. 19A and 19C) or 64 g/L sorbate (FIGs. 19B and 19D), polished by an edge die (FIGs. 19A and 19B) or a center die (FIGs. 19C and 19D), demonstrating the surface finish and polishing uniformity as a function of the sorbate concentration;
FIGs. 20A-D present SEM images of 90 nm copper wiring, showing top view (FIGs. 20A and 20B) and side view (FIGs. 20C and 20D), polished with CMP slurries having 4 g/L sorbate (FIGs. 20A and 20C) or 64 g/L sorbate (FIGs. 20B and 20D), wherein copper corrosion sites are marked by arrows in FIG. 20C;
FIG. 21 presents the dishing values of copper van der-Pauw (VDP) structures
(80 x 80 μιη2) plotted as a function of sorbate concentration in the slurry;
FIG. 22 presents a box plot of normalized copper sheet resistance values at different copper densities (PD) as a function of sorbate concentrations, wherein the sheet resistance values are normalized to a maximum value obtained in sample that was polished with a slurry having a sorbate concentration of 4 g/L;
FIG. 23 presents a box plot of normalized copper sheet resistances values as a function of sorbate concentration in the slurry measured from different line widths, wherein the sheet resistance values are normalized to a maximum value obtained from sample polished with a slurry having a sorbate concentration of 4 g/L;
FIG. 24 summarizes the CMP process with respect to sorbate concentration in a
5 percent by volume H202-glycine based slurry as concluded from the experimental results presented hereinabove. FIGs. 24A-D are schematic diagrams depicting a theoretical surface of a copper sample polished by slurries having two sorbate concentrations, 4 g/L (low sorbate, FIGs. 24A and 24C) and 64 g L (high sorbate, FIGs. 24B and 24D), in two polishing stages, the main CMP process (FIGs. 24A and 24B) and the over-polishing stage (FIGs. 24C and 24D); and
FIGs. 25A-D are electron micrographs at a magnification of 8,000X (FIG. 25A),
64,000X (FIG. 25B), ΙΟ,ΟΟΟΧ (FIG. 25C) and 50,000X (FIG. 25D) of copper- containing wafer samples subjected to CMP before post-treatment with 0.4 M sorbate (FIGs. 25A and 25B) and after post-treatment with 0.4 M sorbate (FIGs. 25C and 25D), showing significant decrease in the silica particles adhesiveness as a result of the surface passivation effect imparted by sorbate to the copper surface.
DESCRIPTION OF SPECIFIC EMBODIMENTS OF THE INVENTION
The present invention, in some embodiments thereof, relates to chemical- mechanical planarization (CMP) and, more particularly, but not exclusively, to compositions that are useful for chemical-mechanical planarization (CMP) of copper- containing surfaces, to processes utilizing same and to products produced by these processes.
The principles and operation of some embodiments of the present invention may be better understood with reference to the figures and accompanying descriptions.
Before explaining at least one embodiment of the invention in detail, it is to be understood that the invention is not necessarily limited in its application to the details set forth in the following description or exemplified by the Examples. The invention is capable of other embodiments or of being practiced or carried out in various ways.
As discussed hereinabove, chemical mechanical planarization (CMP) process is a dynamic process which involved a dynamically re-occurring (recursive) and simultaneous chemical corrosion (oxidation), mechanical removal (abrasive) and chemical removal (complexing) and rapid chemical re-passivation (corrosion inhibition) processes. CMP is widely used for producing copper surface with desired characteristics for production of, for example, device wafers and integrated circuits. A successful CMP process depends, at least in part, on the chemical reagents used in the process, or, in other words, on the chemical reagents composing a CMP slurry composition. The corrosion and passivation of copper during CMP is substantially different than static corrosion or static passivation of copper, since the latter do not involve these reactions in a concurrent situation, and further do not occur with recursive mechanical removal of surface layers.
A successful CMP slurry composition should perform so as to balance between removal of copper and passivation of the exposed copper surface. Such a balance is achieved by selecting an oxidizing agent, a complexing agent and a copper-corrosion inhibitor that together perform so as to afford a chemically stable (passivated) and smoothly planarized copper surface.
Abelev, E. et al. [Electrochem. Acta, 52 (2007), p 5150] have disclosed that enhanced copper passivity can be obtained using potassium sorbate-based CMP solutions. It has been shown that the addition of 10 grams/L potassium sorbate (about 0.07 M) to a 1 gram/L potassium sulfate solution provides an extended copper passivity range, up to 1 VAg Agci- In the aforementioned publication, de-passivation and re- passivation characteristics of the protective layers formed at the copper surface in K- sorbate based solutions in the potential region from open-circuit potential (OCP) up to 0.6 V Ag AgCi were studied by the use of a slurry jet instrument. It was shown that the copper re-passivation rate in 10 grams/L sorbate based solutions was fast in a wide potential range. Imaj values were recorded to be less than 1 μΑ, with a linear dependence on the applied potential. It has also been shown that an increase in K- sorbate concentration leads to an increase in lmax values, and that an increase in sorbate concentration from 2 to 10 grams/L did not lead to a significant change in values, while the introduction of higher K-sorbate caused a considerable increase in the current values. In that study, the increase in lmax values was attributed to the increase in the solution's conductivity. This study, while suggesting sorbate as a static corrosion inhibitor for copper, did not demonstrate its use in a dynamic process such as the CMP process. Moreover, further studies discussed hereinbelow, showed that an increase in sorbate concentrations exhibits an adverse effect on copper etching rates, namely showed an increase of the etching rate instead of inhibition thereof.
Thus, Prasad, Y.N. et al. [Electrochimica Acta, 52 (2007), pp. 6353-6358] have studied the effect of various copper-etching inhibitors on the static etch rate of copper, when added to a CMP slurry composition containing hydrogen peroxide as an oxidizer (typically at a concentration of 5 weight percents) and arginine as an alkaline complexing agent. In Prasad, Y.N. et al. it has been shown that under the tested conditions, potassium sorbate, at a concentration of 0.1 M, moderately decreased the static etching rate, whereby a higher concentration of 0.5 M sorbate resulted in increased static etching rate, thus indicating an "antagonistic" effect of the described combination at such sorbate concentrations. Prasad, Y.N. et al. have not discussed any other effect of the compositions taught therein on copper surface properties such as dishing, smoothness/roughness of the surface or the chemical composition of the planarized surface.
The present inventors have recognized that an effective performance of sorbate in the dynamic CMP process depends, at least in part, on the sorbate concentration, as well as on type and/or concentration of the other components in the CMP composition, as is further supported by the above-described teachings.
While continuing to investigate the performance of sorbate-containing copper CMP slurry compositions in order to obtain an optimal formulation, the present inventors have developed methodologies that allowed following their performance at high sorbate concentrations, using measurements of various direct and indirect CMP performance parameters, particularly parameters which stem from the dynamic CMP process, such as rate of copper removal, dishing, polishing uniformity, smoothness/roughness and re-passivation kinetics.
While successfully reducing their new approach for following CMP performance to practice, the present inventors have surprisingly uncovered that the full potential of sorbate-containing copper CMP slurry compositions becomes apparent at sorbate concentrations higher than 0.2 M (higher than about 30 grams of potassium sorbate per liter aqueous solution, namely higher than about 30 grams/L K-sorbate), and higher even than 0.4 M sorbate anion concentrations. Such findings could not have been attainable using the measurements of Ι^ and other parameters of static copper electrochemistry, as the CMP process is a dynamic process, involving mechanical aspects as well as a repetitive de-passivation (exposure) and re-passivation (protection) of the copper surface.
The present inventors have further studied the effect of the other factors of the
CMP composition, such as the oxidizing agent and the complexing agent, and have outlined some features that are required for an optimal and well-balanced copper CMP slurry composition which affords simultaneously a high copper removal rate and a rapid and stable re-passivated copper surface during and after the CMP process.
Hence, according to an aspect of some embodiments of the present invention, there is provided a copper CMP composition which includes:
a peroxide in a concentration that ranges from 0.1 percent by volume to 10 percent by volume;
a complexing agent in a concentration that ranges from 1 gram per liter to 40 grams per liter; and
a sorbate anion in a concentration of at least 0.2 mol per liter.
As used herein, a "CMP composition" describes a composition useful in a chemical-mechanical planarization process, as this term is widely understood in the art. A CMP composition is thus comprised from abrasive particles for effecting mechanical planarization and from chemical reagents for effecting chemical processes. Accordingly, a CMP composition is often in a form of a slurry and is referred to herein and in the art also as "CMP slurry composition" or simply as "CMP slurry".
As used herein a "copper CMP composition" describes a CMP composition that is suitable for chemical-mechanical planarization of copper-containing surfaces.
By "copper-containing surfaces" it is meant a substance that has a surface, and wherein the surface comprises at least 50 weight percents copper, at least 60 weight percents copper, at least 70 weight percents copper, at least 80 weight percents copper, at least 90 weight percents copper, at least 95 weight percents copper, or being 100 percents copper.
Examples of substances having copper-containing surfaces are delineated hereinbelow, in the context of the disclosed CMP process.
A copper CMP composition typically includes an oxidizer that is suitable for effecting copper corrosion and a complexing agent that is suitable for complexing copper ions in the formed copper oxides, and thus for dissolving copper oxides. The substance may be made, at least in part, from copper, or can include copper only at its surface, as a coating. The copper can be found on the surface of the substrate in any shape and form, such as dots, lines, polygonal or circular shaped areas and any combination thereof. Herein throughout, "molar", also denoted as "M", refers to an absolute concentration of the described substance in the CMP composition, namely, by mols per liter composition.
By "grams per liter" it is meant a weight of the substance per 1 liter of the composition.
According to some embodiments of the present invention, the CMP slurry composition is essentially aqueous, as the process is effected in a conducting medium such as water.
When using a sorbate anion source such as potassium sorbate, the concentration of the sorbate anion can be expressed in grams per liter potassium sorbate, hence concentrations higher than 0.2 mols per liter correspond to concentrations of potassium sorbate (150.22 grams per mol) higher than about 30 grams/L.
Other sources of sorbate anion may include, without limitation, sodium sorbate and other soluble salts of sorbic acid.
Without being bound by any particular theory, the present inventors have suggested that the role of the sorbate seems to negate the role of the other two components in the CMP composition (the oxidizing agent and the complexing agent), as the sorbate associates with both freshly exposed copper and freshly formed copper oxides surface and protects the copper from being further corroded (passivation and re- passivation role), while the oxidizing agent oxidizes the copper to various copper- oxygen species which are more soluble and less resistant to mechanical polishing, and copper complexing agent assists the copper removal process by tipping the equilibrium further towards copper dissolution.
As can be seen in the Examples section that follows below, it is an altogether non-trivial task to find an effective balance between the negating requirements of removing copper atoms from the surface of the substrate, and keeping the freshly exposed copper surface passivated and hence protected from uncontrolled corrosive processes that introduce defects in the fine copper features which are being produced in the substrate.
Specifically, as can be seen in Example 2 hereinbelow, when following the parameter of copper removal rate from blanket wafers, optimal performance of copper planarization and passivation is obtained at sorbate concentrations higher than 0.2 M sorbate (30 grams/L K-sorbate) under conditions of high concentrations of H202, indicating that the corrosive effect of the oxidizer has reached saturation, and the protective effect of the sorbate becomes apparent by lowering the rate of copper removal (see, for example, Table 2 and Table 3 hereinbelow).
Following other parameters of the CMP process, such as X-ray photoelectron spectroscopy of the planarized copper surface (see, Example 3), static contact angles of de-ionized water at the planarized copper surface (see, Example 4) and atomic force microscopy measurements of the planarized copper surface (see, Example 6), it became apparent that the higher the K-sorbate concentration the better performance of a CMP slurry containing same is achieved. Thus, for example, CMP compositions that contain from 16 grams per liter (about 0.1 M sorbate anion) to 64 grams per liter (about 0.4 M sorbate anion) perform better than compositions containing 4 grams per liter (about 0.025 M sorbate anion). It is noted briefly that XPS measurements follow the chemical composition of the sorbate-passivated copper-containing substrate surface at various angles (corresponding to probing the surface at various depths), contact angle measurements quantify the change in the surface tension of a copper-containing substrate as a result of passivation with sorbate, and AFM measurements quantify the smoothness of the planarized copper surface as a function of the sorbate content in the CMP slurry.
Other measurements conducted on copper sheets, processed using copper CMP slurries according to embodiments of the present invention, such as sheet resistance values at different patterned copper densities, indicated that CMP performance can be improved even further at sorbate concentrations higher than 0.4 M, namely at a range of about 0.4 M sorbate anion to about 1.7 M sorbate anion (see, Example 9 hereinbelow).
Hence, according to some embodiments of the present invention, the sorbate anion concentration in the copper CMP slurry is higher than 0.2 M, higher than 0.4 M, higher than 0.5 M, higher than 1.7 M or higher than 2 M. In terms of grams per liter of K-sorbate, its concentration in the copper CMP slurry is higher than 30 grams/L, higher than 40 grams/L, higher than 50 grams/L. higher than 60 grams/L, higher than 64 grams/L, higher than 70 grams/L, higher than 80 grams/L, higher than 90 grams/1, higher than 100 grams/L, and even higher than 200 grams/1 or higher than 256 grams/L. Thus, in some embodiments, a concentration of a sorbate anion in the CMP slurry can be, for example, 0.200 M, 0.206 M, 0.213 M, 0.220 M, 0.226 M, 0.233 M, 0.240 M, 0.246 M, 0.253 M, 0.260 M, 0.266 M, 0.273 M, 0.280 M, 0.286 M, 0.293 M, 0.300 M, 0.306 M, 0.313 M, 0.320 M, 0.326 M, 0.333 M, 0.340 M, 0.346 M, 0.353 M, 0.359 M, 0.366 M, 0.373 M, 0.379 M, 0.386 M, 0.393 M, 0.399 M, 0.406 M, 0.413 M, 0.419 M, 0.426 M, 0.433 M, 0.439 M, 0.446 M, 0.453 M, 0.459 M, 0.466 M, 0.473 M, 0.479 M, 0.486 M, 0.493 M, 0.499 M, 0.506 M, 0.513 M, 0.519 M, 0.526 M, 0.533 M, 0.539 M, 0.546 M, 0.553 M, 0.559 M, 0.566 M, 0.572 M, 0.579 M, 0.586 M, 0.592 M, 0.599 M, 0.606 M, 0.612 M, 0.619 M, 0.626 M, 0.632 M, 0.639 M, 0.646 M, 0.652 M, 0.659 M, 0.666 M, and even higher concentrations such as from 0.666 M to 1.664 M, and even up to about 2 M and higher, which correspond, respectively to a potassium sorbate concentration of 30 grams/L, 31 grams/L, 32 grams/L, 33 grams/L, 34 grams/L, 35 grams/L, 36 grams/L,, 37 grams/L, 38 grams/L, 39 grams/L, 40 grams/L, 41 grams/L, 42 grams/L, 43 grams/L, 44 grams/L, 45 grams/L, 46 grams/L, 47 grams/L, 48 grams/L, 49 grams/L, 50 grams/L, and preferably higher, namely, 51 grams/L, 52 grams/L, 53 grams/L, 54 grams/L, 55 grams/L, 56 grams/L, 57 grams/L, 58 grams/L, 59 grams/L, 60 grams/L, 61 grams/L, 62 grams/L, 63 grams/L, 64 grams/L, 65 grams/L, 66 grams/L, 67 grams/L, 68 grams/L, 69 grams/L, 70 grams/L, 71 grams/L, 72 grams/L, 73 grams/L, 74 grams/L, 75 grams/L, 76 grams/L, 77 grams/L, 78 grams/L, 79 grams/L, 80 grams/L, 81 grams/L, 82 grams/L, 83 grams/L, 84 grams/L, 85 grams/L, 86 grams/L, 87 grams/L, 88 grams/L, 89 grams/L, 90 grams/L, 91 grams/L, 92 grams/L, 93 grams/L, 94 grams/L, 95 grams/L, 96 grams/L, 97 grams/L, 98 grams/L, 99 grams/L, 100 grams/L, and even higher concentrations such as from 100 to 250 grams/L, and even up to 300 grams/L and higher.
It is noted herein that sorbate is characterized by having an exceptional high solubility in water (9 M), which allows using high concentrations in a CMP aqueous composition. It is further noted that as discussed herein and is further demonstrated in the Examples section that follows, such high concentrations of sorbate are required to achieve the balance between rapid corrosion versus rapid re-passivation in a dynamic CMP process, and further achieve desired characteristics of the obtained planarized copper surface, probably due to the sorbate rapid adsorption to copper. It can therefore be deduced that substances that may impart similar adsorption to planarized copper surface but which do not exhibit such a high solubility, and thus cannot be used at high concentrations as described herein, are less likely to exhibit beneficial performance as a copper-corrosion inhibitor in CMP compositions as described herein.
As discussed herein, balancing the requirements for a high copper removal rate and rapid re-passivation of the freshly polished copper surface, require balancing the rates of copper passivation, effected by the sorbate anion, versus copper corrosion, effected by the oxidizing agent and the complexing agent. Specifically, the role of an oxidizing agent such as hydrogen peroxide has been identified as providing a layer of copper oxides/hydroxides species at varying rates, which depend on the H202 concentration in the CMP composition (see, the Examples section that follows). Without being bound by any particular theory, it is assumed that a high concentration of the oxidizing agent leads to a faster rate of formation of copper oxides/hydroxides and to a thicker layer of oxides. While faster and deeper copper oxidation may lead to faster copper removal rates, it may also lead to more dishing and an enhanced roughness of the obtained planarized copper surface.
It is noted herein that the corrosive reaction which is required for the CMP process, can be achieved by various oxidizing agents other than a peroxide.
Hence, according to some embodiments of the present invention, a composition of the present invention may include any type of copper oxidizing agents, including, for example, peroxides, phenols, permanganates, chromates, iodates, iron salts, aluminum salts, sodium salts, potassium salts, phosphonium salts, chlorates, perchlorates, persulfates oxidizing agents, and any mixtures thereof.
Examples of suitable oxidizing agents include, but are not limited to, phenol,
KI03, KBr03, K3Fe(CN)6, K2Cr207, V203, H202, HOC1, KOC1, KMg04 and KMn04.
According to some embodiments of the present invention, a peroxide is used as an oxidizing agent capable of oxidizing copper to copper oxide/dioxide. As used herein, the term "peroxide" is a chemical substance containing the peroxide anion (02 ~) or a single oxygen-oxygen bond, and capable of forming, even as an intermediate, a reactive oxygen radical species. Peroxides include both inorganic peroxides and organic peroxides. Non-limiting examples of inorganic peroxides include hydrogen peroxide, sodium peroxide, barium peroxide, calcium peroxide and magnesium peroxide. Organic peroxides include, without limitation, benzoyl peroxide and methyl ethyl ketone peroxide.
The peroxide is used in a concentration that ranges from 5 percent by volume to
10 percent by volume.
According to some embodiments of the present invention, the peroxide is hydrogen peroxide (H202).
As discussed hereinabove, while the oxidizing agent promotes copper oxidation, the complexing agent pushes the equilibrium towards copper dissolution by forming complexes with various copper ions which form as a result of the oxidation reaction.
According to some embodiments of the present invention, the complexing agent can be any copper complexing agent known in the art which is suitable for use in neutral and alkaline (basic) slurries.
According to some embodiments of the invention, a suitable copper complexing agent is such that, in combination with the high sorbate concentrations disclosed herein and the peroxide oxidizing agent, performs such that the required balance between copper removal rate (copper dissolution) and copper passivation is achieved under dynamic CMP conditions.
A skilled artisan would find ample guidance to methods and criteria for selecting a suitable complexing agent in the Examples section that follows below, particularly, but not exclusively, in the techniques and methods presented in Examples 2, 5, 7, 8, 9 and 10 hereinbelow.
In the context of some embodiments of the present invention, a suitable copper complexing agent is selected such that the increase in the corrosion inhibition activity of sorbate exhibited at increased sorbate concentrations is not reversed or flattened in the presence of the complexing agent, and vice versa, that the inhibitory effect of sorbate does not diminish the effective copper dissolution enhancement activity of the complexing agent. Non-suitable complexing agents, according to some embodiments of the present invention, are those with which the above-described performance of sorbate is reversed or flattened, and/or those that cannot effect copper complexation effectively in high sorbate concentrations. According to some embodiments of the present invention, suitable complexing agents include, without limitation, amine-containing agents, such as, but not limited to, amino acids, monoamine-containing compounds and diamine-containing compounds. Exemplary diamines with aliphatic linear carbon chain include compounds such as ethylene diamine (1,2-diaminoethane), 1,3-diaminopropane (propane-l,3-diamine), putrescine (butane- 1,4-diamine), cadaverine (pentane-l,5-diamine) and hexamethylenediamine (hexane-l,6-diamine). Other exemplary diamines include ethambutol, dimethyl-4-phenylenediamine, N,N'-di-2-butyl-l,4-phenylenediamine, diphenylethylenediamine and 1,8-diaminonaphthalene. Other exemplary suitable complexing agents include glycolic acid, phthalic acid salt, citric acid, and oxalic acid.
Amino acids which are suitable complexing agents include, but are not limited to, glycine, arginine, phenyl alanine and serine.
As noted hereinabove, a performance of a CMP composition which contains a high concentration of sorbate can be attenuated by the concentration of the complexing agent used. In order to achieve a desired performance of the composition, according to some embodiments of the invention, the concentration of the complexing agent in the CMP composition ranges from 0.07 mol per liter (M) to 0.5 mol per liter (M), or from 0.1 M to 0.5 M. Alternatively, the concentration of the complexing agent in the CMP composition is about 0.07 M, about 0.1 M, about 0.2 M, about 0.3 M, about 0.5 M, or about 0.5 M.
As stated herein, Prasad, Y.N. et al. used arginine as a complexing agent with sorbate, but have failed to show a beneficial effect of combining the two agents, presumably since Prasad, Y.N. et al. have used non-dynamic criteria to assess their compositions; hence, showing that the etch rate increased with an increase of sorbate concentration while studying arginine-containing slurry compositions, probably due to changes in the electrical conductance.
According to some embodiments of the present invention, the complexing agent is glycine.
Glycine is the simplest amino acid, represented as NH2-CH2-C(=0)OH. In water, it has a PH 7, and a neutral charge. In acidic solution, it is present in a form of an ammonium salt, whereby in basic solutions, it is present as a carboxylate anion. CMP composition that contain glycine are expected to act differently than CMP composition that contain "basic" complexing agents such as arginine.
As demonstrated in the Examples section that follows, a CMP composition exhibiting high performance was obtained with high sorbate concentration using hydrogen peroxide as an oxidizing agent and glycine as a copper complexing agent.
Hence, according to an aspect of some embodiments of the present invention, there is provided a copper CMP composition which includes hydrogen peroxide as an oxidizing agent, glycine as a complexing agent and a sorbate anion at a concentration higher than 0.2 mol per liter.
Without being bound by any particular theory, it is noted herein that hydrogen peroxide and glycine cross-enhance their reactivity throughout a catalytic mechanism in which H202 decomposes by Cu-glycine complex to yield hydroxyl radicals. The former is assumed to be a more powerful oxidizer than H202. Thus, more copper is oxidized, resulting in more copper ions available to complex with glycine. The copper removal rate in either H202 or glycine solution alone is not significant while when the two present together, high removal rates can be achieved.
Hence, according to some embodiments of this aspect of the present invention, the concentration of hydrogen peroxide in the copper CMP slurry composition may range from 0.1 percent by volume to 10 percents by volume. Alternatively, as demonstrated in the Examples section that follows, the concentration of H202 may range from 0.3 percent by volume to 1 percents by volume (referred to as low H202 concentration), and further alternatively, the concentration of H202 ranges from 5 percent by volume to 10 percents by volume (referred to as high H202 concentration) (see, e.g., Example 5 hereinbelow).
According to some embodiments of this aspect of the present invention, the concentration of glycine in the copper CMP slurry composition may range from 5 grams per liter to 40 grams per liter. Alternatively, as demonstrated in the Examples section that follows, the concentration of glycine ranges from 5 grams per liter to 10 grams per liter (see, e.g., Example 2 hereinbelow).
Hence, according to some embodiments of the present invention, the copper
CMP slurry composition includes sorbate anion in a concentration that ranges from 0.2 mols per liter to 9 mols per liter, glycine in a concentration of 10 grams per liter (about 0.07 M), and hydrogen peroxide in a concentration that ranges from 5 percent by volume to 10 percents by volume.
As presented hereinabove, most known copper CMP slurry compositions which are acidic pose difficulties stemming mostly from uncontrolled effect of the acidity on the machinery and products of the wafer manufacturing process, while basic slurries are less prone to such problems. Hence, according to embodiments of the present invention, the pH of the copper CMP composition ranges from 7 to 14, optionally, from 7 to 10, optionally from 8 to 9, and optionally the pH is about 8.5. Depending of the intrinsic pH of the pre-adjusted solution, the pH can be set to the desired level by adding small amounts of base or acid. According to some embodiments of the present invention, the pH of the otherwise acidic solution of hydrogen peroxide, sorbate and glycine can be adjusted to 8.5 by adding KOH, NaOH or other suitable bases.
Thus, a CMP composition described herein can further include a pH-adjusting agent (e.g., a base, such as an alkali metal hydroxide, for example, KOH or NaOH).
It is noted herein that a suitable complexing agent in the context of embodiments of the present invention, is one that is suitable for alkaline slurry compositions, such as amine-containing complexing agents.
In some embodiments, a copper CMP composition as described herein may further include additional components, including, for example, additional complexing agents, and/or metal salts (e.g., potassium sulfate.
According to some embodiments of the present invention, the composition presented herein further comprises abrasive particles.
The abrasive particles are used to effect mechanical planarization and are selected to selectively remove passivated copper (in the form of copper oxides). Exemplary particles are made of metal oxides. Suitable metal oxides include, but are no limited to, oxides of silicon, aluminum, cerium, germanium, titanium, zirconium and mixtures thereof. Specific examples of suitable abrasives include particles of Si02, Ce02, A1203, SiC, Si3N4 and Fe203.
In some embodiments of the present invention the abrasive particles make up between about 1 % and 30 % by weight of the composition.
According to another aspect of the present invention, there is provided a process of chemical-mechanical planarization of a copper-containing surface which is effected by contacting a substrate having a copper-containing surface with the copper CMP slurry composition as presented herein, thereby planarizing the copper-containing surface. This process is also referred to herein as CMP process.
As discussed hereinabove, a CMP process involves a simultaneous chemical (e.g. by copper-oxidizing and copper-complexing agents) and mechanical (by means of abrasives) planarization of copper-containing surfaces, and a well balanced CMP process also provides a freshly exposed yet passivated copper surface. It has also been discussed hereinabove, that planarization of the bases layers in a multi-level wafer- based device manufacturing is required to ensure a flat surface prior to subsequent layering and/or lithography, and that passivation is required to maintain the planarization though the various steps of the process. Hence, a beneficial CMP process is required to produce a completely flat and smooth surface of the base layer that stays protected from further uncontrolled corrosion.
Accordingly, planarizing a copper-containing surface is defined herein as the removal of protruding or uneven parts of the surface in a copper-containing substrate so as to arrive at a planar and leveled surface at a minimal loss of substance. The CMP process, according to some embodiments of the present invention, would also provide, apart from planarization, a passivated copper surface, namely planarized copper surface which would not undergo further etching or uncontrolled corrosion.
According to some embodiments of the present invention, the process further includes, subsequent to contacting the substrate with the slurry, applying a polishing pad onto the substrate to effect polishing of the copper-containing surface of the substrate.
Faster or slower rates of copper removal using any given CMP slurry, as well as other parameters of the CMP process such as dishing, may also be achieved by mechanical means, namely polishing pad's shape, material, speed, down pressure and angle; however, as discussed herein, the rate of removal of copper from the surface of the substrate at any given mechanical conditions, also depends on the composition of the slurry. Hence, according to some embodiments of the present invention, the process is characterized by a copper removal rate of at least 200 nanometer per minute under a given set of mechanical conditions, such as those presented for example in Table 1 in Example 1 hereinbelow. It is noted that optimal removal rates would be different using identical slurries under different mechanical conditions; hence the rate of copper removal should be regarded as "normalized" against any given set of mechanical conditions.
It is noted herein that seeking to improve the parameter of copper removal rate alone will ultimately not ensure optimal results; while excess oxidizing agent has been shown to increase the rate of copper removal, it may also reduce the overall quality of the CMP process by increasing dishing and other signed of uncontrolled corrosion, leading to rough and un-passivated surface.
As used herein, the term "dishing" refers to the phenomenon of uneven, non- planar surface polishing result, wherein sunk regions are formed typically near the center of an otherwise flat/planar area due to various reasons. In the context of embodiments of the present invention, dishing is associated with uncontrolled corrosion of the surface material, and can be limited or resolved by rapid re-passivation of the surface during the CMP process, which involved dynamic and simultaneous corrosion and re-passivation reactions.
Given all other parameters equal, it is appreciated that dishing also depends on the span and geometry of the polished surface, as small areas will be less prone to dishing compared to wide areas. Hence, parameterization of dishing should refer to mechanical parameters, as in the case of removal rates discussed hereinabove, and to the size and shape of the copper surface being planarized.
Hence, the process, according to some embodiments of the present invention, is characterized by a dishing value of less than 110 nanometer (see, e.g., Example 8 hereinbelow) when measured under a given set of mechanical conditions as presented for example in Table 1 in Example 1 hereinbelow, and given a sample of copper having a van der-Pauw (VDP) structure of 80 μπι x 80 μπι in size.
As the intermediate product of a manufacturing process involving copper CMP is typically a substrate having a planarized and passivated copper-containing surface, there is provided a substrate having a planarized and passivated copper-containing surface obtained by the process presented herein, using the CMP slurry composition presented herein.
While the end-product of manufacturing a wafer-based electronic device may have copper-containing surfaces embedded therein which have been processed by the presently disclosed CMP compositions and processes, these surfaces may be inaccessible due to layers that have been added on top during the formation of a complete electronic device. The quality of the entire process of manufacturing electronic wafer-based devices, which is also measured by the rate of disqualification of malfunctioning devices on a wafer, is however improved by using the compositions and processes disclosed herein.
As demonstrated in the Examples section hereinbelow, AFM measurements of the roughness of the planarized and passivated copper-containing surface showed the effect of sorbate on that product characterization parameter. Hence, according to some embodiments of the present invention, the planarized and passivated copper-containing surface may be characterized by height range (Z range representing roughness/smoothness) value of less than 20 nm or less than 15 nm (see, e.g., Example 6 hereinbelow).
As demonstrated in the Examples section hereinbelow, the presence of a protective (passivating layer) of sorbate on the planarized and passivated copper- containing surface can be determined and even quantified by the shape of a drop of deionized water placed on the planarized and passivated copper-containing surface. Hence, according to some embodiments of the present invention, the copper surface may by further characterized by a water-to-surface contact angle of at least 80 ° (see, e.g., Example 4 hereinbelow).
Another parameter by which a substrate having a planarized copper-containing surface, obtained by the composition and process presented herein can be characterized, is the resistance of a sheet of copper processed thereby. Such copper sheets or patterned copper lines with various line widths can be used to simulate an integrated wafer-based circuit for purposes of evaluating the performance of the CMP process and compositions. As demonstrated in Example 9 hereinbelow, sheet resistance values were obtained from samples treated by the process and compositions presented herein, and normalized against the values obtained from an identical sample processed with a composition containing a low concentration of 0.03 M sorbate (corresponding to about 4 g L K-sorbate).
Hence, according to some embodiments of the present invention, the copper sheet resistance values characterizing a substrate having a planarized copper-containing surface treated by the process and compositions presented herein, and having copper pattern density 10-60 %, ranges from 0.2 arbitrary units (A.U.) to 0.4 A.U.
According to some embodiments of the present invention, the copper sheet resistance values characterizing a substrate having a planarized copper-containing surface treated by the process and compositions presented herein, and having copper lines width of 1-30 μηι, ranges from 0.1 A.U. to 0.3 A.U.
Another criterion for assessing the benefits of a CMP composition is the "cleanness" of the copper surface at the end of the CMP process, namely the absence of remaining abrasive particles thereon after the CMP process.
As demonstrated in Example 10 hereinbelow, the use of high concentrations of sorbate affects the surface so as to substantially prevent adhesion of abrasive particles to the polished surface.
According to some embodiments of the present invention, the substrate having a planarized copper-containing surface forms a part of an article of manufacturing.
In some embodiments, the substrate is integrated in an article as a copper- containing element.
Exemplary articles of manufacturing which may incorporate copper-containing surfaces or copper-containing elements in their final form or in an intermediate prepared during manufacturing the articles include, but are not limited to, an integrated circuit, a wafer-based device, 3D integration and advanced interconnects.
The phrase "wafer-based device", as used herein, referred to a section of a wafer on which a copper-containing electronic device has been produced by using the CMP process and composition presented herein among other processes and manufacturing steps. Hence, according to another aspect of embodiments of the present invention, there is provided an article which includes a wafer-based device having a copper- containing element, wherein the copper-containing element is being produced by the process presented herein; the article is characterized by a copper dishing value of less than 110 nanometer or less than 50 nm and a height range (Z range) value of less than 20 nm or less than 15 nm.
A "copper-containing element" describes an element in an article that comprises at least 50 weight percents copper, at least 60 weight percents, at least 70 weight percents, at least 80 weight percents, at least 90 weight percents copper, or is made of 100 percents copper. The element is the final product is obtained upon planarizing an intermediate that had a copper-containing surface. Alternatively, the final product, the article, has a copper-containing surface.
The CMP slurry composition, according to some embodiments of the present invention, embodies and incorporates a balanced combination of two opposing chemical processes, copper removal by corrosion and aberration, versus rapid copper passivation to corrosion. This non-trivial balance is achieved by specific concentrations of agents which accomplish the chemical processes, namely oxidizers such as H202 and complexing agents such as glycine, opposing sorbate as a passivating/inhibiting/re- passivating agent. The effect of this non-trivial balance can be measured by the rate of copper removal, indicating an effective corrosion process; a low dishing rate, indicative of the rapid re-passivation of the passivation process; low Z-range of the surface, indicating smoothness; and low resistance of test copper sheets, indicating an overall successful CMP process, which ultimately predicts and ensures a low rate of disqualified end-product devices on a wafer.
It is expected that during the life of a patent maturing from this application many relevant high sorbate-content copper CMP slurry compositions will be developed and the scope of the term high sorbate-content copper CMP slurry composition is intended to include all such new technologies a priori.
As used herein the term "about" refers to ± 10 %.
The terms "comprises", "comprising", "includes", "including", "having" and their conjugates mean "including but not limited to".
The term "consisting of means "including and limited to".
The term "consisting essentially of" means that the composition, method or structure may include additional ingredients, steps and/or parts, but only if the additional ingredients, steps and/or parts do not materially alter the basic and novel characteristics of the claimed composition, method or structure.
The word "exemplary" is used herein to mean "serving as an example, instance or illustration". Any embodiment described as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments and/or to exclude the incorporation of features from other embodiments. The word "optionally" is used herein to mean "is provided in some embodiments and not provided in other embodiments". Any particular embodiment of the invention may include a plurality of "optional" features unless such features conflict.
As used herein, the singular form "a", "an" and "the" include plural references unless the context clearly dictates otherwise. For example, the term "a compound" or "at least one compound" may include a plurality of compounds, including mixtures thereof.
Throughout this application, various embodiments of this invention may be presented in a range format. It should be understood that the description in range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the invention. Accordingly, the description of a range should be considered to have specifically disclosed all the possible subranges as well as individual numerical values within that range. For example, description of a range such as from 1 to 6 should be considered to have specifically disclosed subranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6 etc., as well as individual numbers within that range, for example, 1, 2, 3, 4, 5, and 6. This applies regardless of the breadth of the range.
Whenever a numerical range is indicated herein, it is meant to include any cited numeral (fractional or integral) within the indicated range. The phrases "ranging/ranges between" a first indicate number and a second indicate number and "ranging/ranges from" a first indicate number "to" a second indicate number are used herein interchangeably and are meant to include the first and second indicated numbers and all the fractional and integral numerals therebetween.
As used herein the term "method" refers to manners, means, techniques and procedures for accomplishing a given task including, but not limited to, those manners, means, techniques and procedures either known to, or readily developed from known manners, means, techniques and procedures by practitioners of the chemical, pharmacological, biological, biochemical and medical arts. It is appreciated that certain features of the invention, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the invention, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable subcombination or as suitable in any other described embodiment of the invention. Certain features described in the context of various embodiments are not to be considered essential features of those embodiments, unless the embodiment is inoperative without those elements.
Various embodiments and aspects of the present invention as delineated hereinabove and as claimed in the claims section below find experimental support in the following examples. EXAMPLES
Reference is now made to the following examples, which together with the above descriptions; illustrate the invention in a non limiting fashion.
MA TERIALS AND METHODS
Electrochemical studies:
A pencil-type copper electrode, (5 mm diameter rod, 99.9995 percent by weight) mounted in a room temperature-curing-epoxy, was used subsequent to polishing (800 grit SiC paper). Studies were performed in a three-electrode electrochemical cell with the use of a potentiostate (273A EG&G).
Working electrode potentials were referred to a saturated calomel electrode
(SCE, 0.245 VSHE) which was connected through a Luggin-Habber capillary tip assembly, while the counter electrode was a Pt plate.
Potentiodynamic and potentiostatic studies were carried out subsequent to a cathodic pretreatment, in which the copper electrode was exposed to a potential of -1.3 V(scE) for 1 minute in order to reduce pristine copper oxides covering the electrode. Potentiodynamic characteristics of copper electrodes in different electrolytes were obtained at a scan rate of 5 mVs"1.
Cyclic polarization curves were obtained by reversing the scan direction once the recorded anodic potential reached a value of 1 V. Polishing Tool and Set-up:
CMP measurements were performed on 200 mm blanket wafers (E-460 Mecapol polisher, Alpsitec). Consumables and process parameters in the polishing experiments are summarized in Table 1.
Table 1
Figure imgf000032_0001
CMP of patterned wafers was carried out on an Applied Materials Mirra Mesa polishing tool for more accurate polishing. CMP was performed with 25 % over- polishing time.
CMP slurries:
Slurries were prepared with de-ionized water (DI, 18 ΜΩ-cm, Millipore system) with the addition of analytical grade chemicals; potassium sorbate, (CH3(CH)4COO~K+), glycine (H2NCH2COOH) and hydrogen peroxide (30 percents by weight H202). The pH of the copper CMP slurries was adjusted to 8.5 with addition of KOH. Barrier CMP was performed with a commercial slurry with pH value of 2.6.
Contact Angle Measurements:
Static contact angles of de-ionized water at the copper surface were measured in air with an OCA-20 contact angle system. X-ray photoelectron spectroscopy (XPS) Measurements:
Surface modification of copper was characterized by using X-ray photoelectron spectroscopy (XPS, a theta 300 Mkll system (Thermofisher), having Al -Ka X-radiation (1486.6 eV)). The in-depth film composition and distribution of the oxidation state across the protective film on the copper surface were determined by an Angle Resolved XPS technique (ARXPS). The experiments were conducted at take-off angles ranging from 69° (bulk mode) to 12° (surface mode) with respect to the copper surface.
High resolution profilometer (HRP):
A KLA-Tencor high resolution profilometry tool was used to scan various structures and arrays on the patterned wafers in order to construct the relative surface topography and evaluate dishing values. Dishing values were measured on 80 μπϊ Van Der Pauw (VDP) structures.
Four probe point tool for sheet resistance measurements:
The measurements were performed using Prometrix OmniMap RS75 four probe system. Sheet resistance was measured as follows:
Measured X Wire width = Rh0
wire length wire height
measured Rri,„,
normalized r¾
where, Rh0 is the resistivity of the sample. Sheet resistance values were normalized to a maximum value obtained in sample that was polished with low sorbate concentration.
Atomic force microscopy (AFM):
AFM measurements were conducted using tapping mode with Nanoscope IV- Vecco AFM. EXAMPLE 1
GENERAL SLURRY CHEMISTRY
In order to understand the individual contribution of each slurry component on copper behavior, anodic polarizations of copper in solutions containing different concentrations of each component were performed.
Copper polarization studies in K-sorbate solutions:
Figure 1 is a comparative plot showing 2 anodic polarization curves of copper obtained by polarizing a copper electrode in a solution containing one gram per liter (g/L) of potassium sulfate (1 g/L K2S04, curve denoted 'a'), and from polarizing copper in a solution containing one gram per liter of potassium sulfate and 8 grams per liter of potassium sorbate (curve denoted b").
As can be seen in Figure 1, copper polarization in the sulfate solution exhibits an active dissolution behavior, with no evidence of passivation, which is a typical active anodic behavior of copper in a sulfate solution, while in the solution containing potassium sorbate (K-sorbate), low current densities were observed over a wide potential range. As can further be seen in Figure 1, no breakdown of the protective film was observed in the anodic region as the anodic currents remained constant up to a potential of 1 V (SCE)- During the reversed scan, decreased current densities and higher final corrosion potential were observed. This behavior indicates the high stability of the sorbate protective film, displaying a nobler new Ecorr.
Figure 2 is a comparative plot showing three anodic polarization curves of copper obtained by polarizing a copper electrode in a solution containing 4 g/L potassium sorbate (K-sorbate), each further containing either 1 g/L of potassium sulfate (K-sulfate), 1 g/L glycine, or 1 g/L potassium acetate (K-acetate).
As can be seen in Figure 2, in the case of the solution containing 1 g/L glycine the Ecorr was observed at a lower potential (less noble) value than in either acetate or sulfate solutions. As can further be seen in Figure 2, the Icorr recorded in the glycine- containing solution was higher than in either the acetate or sulfate solutions. This observation implies that the anodic dissolution reaction is enhanced in the glycine solution, resulting in a EcorT shift to a less noble potential value. It can further be seen in Figure 2 that in the sulfate and glycine solutions, a concentration of 4 g/L K-sorbate was insufficient to produce a protective film at the copper surface. Thus, a breakdown of the passive layer was observed at potential values of 0.38 and 0.28 V(SCE), respectively. However, in the acetate solution, a concentration of 4 g/L K-sorbate was sufficient enough to produce a protective film at the copper surface and no breakdown was observed up to a potential of 1 V(SCE)- Furthermore, it can be seen that the Icorr obtained in sulfate and acetate solutions containing 4 g/L of sorbate was in the order of 10"6 A/cm2, while in the glycine solution, the Icorr was higher by two orders of magnitude (10"4 A/cm2).
These results suggest that the formation and stability of the protective film are affected by the media it is formed at. Therefore, the inhibition efficiency of sorbate in peroxide (oxidant) and glycine (complexing agent) containing solutions was further studied, as presented hereinbelow.
Copper polarization studies in hydrogen peroxide-glycine solutions:
It should be noted that the electrochemical behavior of copper and the nature of the oxide formed in either H202 or glycine solutions differ significantly from those obtained in solutions containing both H202 and glycine.
Figure 3 is a comparative plot showing 4 anodic polarization curves of copper obtained by polarizing a copper electrode in a solution containing 1 g/L Na2S04 and 10 g/L glycine, and containing no hydrogen peroxide (curve denoted Ί1), 0.3 percents by volume hydrogen peroxide (curve denoted '2"), 1 percents by volume hydrogen peroxide (curve denoted '31), and 5 percents by volume hydrogen peroxide (curve denoted '4').
As can be seen in Figure 3, an increased H202 concentration gradually increased the Ecorr- The addition of 0.3 percent by volume H202 to the glycine solution increased the Icorr, while further increase in the H202 concentration led to a decrease in the Icon- This observation coincides with the trend observed in copper removal rate values which is discussed hereinbelow.
Without being bound by any particular theory, the decrease in the ICorr rnay be attributed to the adsorption of hydroxyl radicals at the copper surface. The radicals are formed due to the decomposition of H202 catalyzed by Cu-glycine complexes. Thus, as can further be seen in Figure 3, the increase in the radicals' concentration gradually expanded the passivity region from a value of 0 mV(scE) (in 0.3 percent by volume H202) to 75 mV(scE) (in 1 percent by volume H202) and 137 mV(scE) (in 5 percent by volume H202).
These observations may indicate a better characteristic of the protective oxide/hydroxide layer which may be attributed to a higher coverage at the copper surface.
In order to observe the progress in the formation of copper oxides/hydroxides with increased H202 concentration, the anodic behavior of copper in solutions containing 1 g/L sodium sulfate, 10 g L glycine and different concentrations of H202 was also studied potentiostatically. The measurements were conducted at a potential value of 0.3 V(SCE) for 200 seconds subsequent to a cathodic pretreatment.
Figures 4A-F present high resolution SEM micrographs (Figures 4A, 4C and 4E, left column) alongside with plots presenting potentiostatic measurements (Figures 4B, 4D and 4F respectively, right column), obtained from a copper electrode polarized at 0.3 V(scE) for 200 seconds in a solution containing 1 g/L Na2S04, and lOg/L glycine and containing 0.3 percents by volume hydrogen peroxide (Figures 4A-B), 1 percents by volume hydrogen peroxide (Figures 4C-D), or 5 percents by volume hydrogen peroxide (Figures 4E-F).
As can be seen in Figures 4A-F, the surface morphology correlates with the current densities recorded from the potentiostatic measurements, wherein the solution containing 0.3 percent by volume H202, the copper surface was hardly covered with copper oxides/hydroxides. Pits (appearing as black spots in the SEM micrograph Figures 4A and 4C) were observed at the copper surface, indicating a local dissolution of the copper. In the solution containing 1 percent by volume H202, the copper surface was covered with copper oxides/hydroxides, which appeared as flakes structures. Interestingly, these structures did not cover the copper surface completely and pits caused by dissolved copper were seen at the surface. As the peroxide concentration increased, the flaky structures increased in size and covered larger portions of the copper surface. Accordingly, it can be seen in Figures 4A-F that an increase in the H202 concentration resulted in decreased current densities, observed in the potentiostatic measurements: about 2.5 mA (in 0.3 percent by volume H202) compared with about 1 mA (in 1 percent by volume H202) and about 0.3 mA (in 5 percent by volume H202). Indeed, the higher coverage of the oxides/hydroxides obtained with the increased H202 concentration may reduce the corrosion rate; however, it cannot offer copper passivation.
These observations assist in understanding the CMP process, since the peroxide concentration in the slurry determines the characteristics of the formed protective film, i.e. its thickness, chemical composition and coverage at the copper surface. These characteristics may eventually determine the possible reactions at the copper surface and consequently the removal rate during the CMP process.
Integrating sulfate, glycine, H2O2 and sorbate:
The electrochemical behavior of copper was studied in solutions containing 1 gram per liter (g L) sodium sulfate and 10 g/L glycine having different compositions of H202 and sorbate. Thus, 0.3, 1 or 5 percent by volume H202 and different concentration of sorbate (from 0 to 64 g/L) solutions were studied.
The three regions of H202 concentrations were selected to illustrate the effect of the sorbate at each region.
Figures 5A-F present comparative plots of potentiodynamic curves (Figures 5A, 5C and 5E, left column) alongside with comparative plots of potentiostatic curves at 300 mV (Figures 5B, 5D and 5F respectively, right column), obtained from a copper electrode polarized in solutions containing 1 g/L Na2S04, 10 g/L glycine and containing 0.3 percents by volume hydrogen peroxide (Figures 5A-B), 1 percents by volume hydrogen peroxide (Figures 5C-D), or 5 percents by volume hydrogen peroxide (Figures 5E-F), and further containing different K-sorbate concentrations as indicated in each plot.
As can be seen in Figures 5A-F, the addition of sorbate to the solution significantly affects the electrochemical behavior of copper. As can be seen in Figure 5A, the onset of the anodic current in the solution containing 0.3 percent by volume H202 was obtained at a potential of 0.05 V(SCE)- A positive potential shift gradually increased the anodic currents further, indicating an active copper dissolution. The onset of the anodic currents with the addition of 4 g/L sorbate was obtained at a potential of 0.06 V(scE)- shift to more positive potential values led to a copper passivity region. At a potential of 0.3 V(SCE), a passivity breakdown was detected; the anodic currents rapidly increased. As can further be seen in Figures 5A-F, a reverse scan at a potential of 1 V(SCE) revealed higher anodic currents. These results indicate that the copper was not fully passivated and suffered from corrosion attacks. While further increasing the sorbate concentration to 16 and 64 g/L, no passivity breakdown was observed up to a potential of 1 V(scE)- However, during the reverse scan, hysteresis was observed in the 16 g/L sorbate solution, while in the 64 g/L sorbate solution, the anodic currents decreased, indicating that the surface maintained its protective layer.
As can be seen in Figure 5C, at 1 percent by volume H202 solution the onset of the anodic currents was obtained at a potential of 0.14 V(SCE)- A more positive potential shift revealed a narrow region of passivity (75 mV). At potentials above 0.2 V(SCE), the anodic currents gradually increased, indicating an active copper dissolution. The addition of 4 g/L sorbate to the solution extended the copper passivity region. However, at a potential of 0.4 V SCE> a breakdown was detected and the anodic currents rapidly increased upon further anodic polarization. The addition of 16 and 64 g/L sorbate to the solution revealed a similar behavior as observed in the 0.3 percent by volume H202 solution.
As can be seen in Figure 5E, at 5 percent by volume H202 solution the onset of the anodic currents was obtained at a potential of 0.2 V(SCE)- A more positive potential shift revealed a narrow region of passivity (137 mV). At potentials above 0.33 V(SCE) the anodic currents gradually increased. The addition of 4 g/L sorbate to the solution revealed a region of copper passivity. At a potential of 0.5 V(SCE) a breakdown was detected and the anodic currents rapidly increased during further positive potential shift. The addition of 16 and 64 g/L sorbate to the solution revealed the same behavior as was observed in the 0.3 and 1 percent by volume H202 solutions.
Potentiostatic measurements performed at a potential value of 300 mV illustrate the stability of the formed protective film over time. In the 0.3 percent by volume H202 solution (Figure 5B), no formation of a protective layer at the copper surface was observed and the anodic currents increased. The addition of 4 g/L sorbate led to a decrease in the anodic currents. However, subsequent to an exposure time of 25 seconds, the anodic currents increased, indicating on the low stability of the protective film, possibly because of a low coverage of the sorbate protective film at the copper surface. In solutions having sorbate concentrations of 16 and 64 g/L, the anodic currents decreased and remained constant during the whole exposure time. This behavior indicates the formation of a highly protective layer at the copper surface.
In the 1 percent by volume H202 solutions (Figure 5D), the anodic currents increased, albeit lower values were obtained compared to the values recorded in the 0.3 percent by volume H202 solution. The addition of 4 g/L sorbate led to a decrease in the anodic currents. However, subsequent to an exposure time of 60 seconds, the anodic currents gradually increased. In solutions having sorbate concentrations of 16 and 64 g/L, the anodic currents decreased and remained constant throughout the time of exposure.
In the 5 percent by volume H202 solutions (Figure 5F), the anodic currents decreased, indicating on the formation of a protective film at the copper surface. The addition of 4 g/L sorbate decreased the anodic currents albeit; subsequent to a 100 seconds exposure time the anodic currents slightly increased. Again, in solutions containing sorbate in concentrations of 16 and 64 g/L, the anodic currents decreased and remained constant during the whole exposure time. These results clearly show that the addition of sorbate to the H202-glycine solutions expands the passivity region in both low and high H202 concentrations. Nevertheless, the enhanced passivity is achieved in the high H202 concentration solution due to the formation of both an oxide and a sorbate protective film at the copper surface.
EXAMPLE 2
CMP STUDIES ON BLANKET WAFERS CMP studies on copper blanket wafers were performed with slurries containing sorbate as an inhibitor, hydrogen peroxide as an oxidizer and glycine as a complexing agent. This study was conducted in order to explore copper dissolution and passivation properties in H202-glycine based slurries containing sorbate as the selected inhibitor. Thus, a balance between the two properties (dissolution-passivation) has been demonstrated. The Effect
Figure imgf000040_0001
and Sorbate on the Copper Removal Rate:
CMP removal rate was studied in slurries containing various concentrations of sorbate (from none to 64 g/L), and peroxide (0.1 to 10 percent by volume), and a fixed amount of glycine (10 g/L).
Figure 6 presents a comparative plot showing the copper removal rate as a function of sorbate content in slurries containing 10 g L glycine with low or intermediate concentrations of H202.
Table 2 presents copper removal rate values as a function of sorbate content in slurries containing 10 g/L glycine with high H202 concentration.
Table 2
Figure imgf000040_0002
As can be seen in Figure 6 and Table 2, the presented data illustrate a typical trend in the copper removal rate as a function of the H202 concentration in the slurry. This trend consists of three regions: (i) low concentrations of H202 (0.1-0.3 percent by volume), (ii) intermediate concentrations of H202 (0.3-1 percent by volume); and (iii) high concentrations of H202 (5-10 percent by volume).
At low concentrations of H202, any addition of H202 led to an increase in the copper removal rate until a maximum value was obtained at a concentration of 0.3 percent by volume H202. At intermediate concentrations of H202, further addition of H202 decreased the copper removal rate up to a concentration of 1 percent by volume H202. Furthermore, it can be seen that the increase in the sorbate concentration from 4 g/L to 64 g/L gradually decreased the removal rates values at both low and intermediate peroxide concentrations.
As can be seen in Table 2, no significant change in the removal rate values was observed at high concentrations of H202 once the H202 concentration increased from 5 to 10 percent by volume. Moreover, the addition of sorbate did not decrease the removal rate in this region, not even at a high concentration of 64 g/L.
It is known that the removal of Cu(I) by both chemical and mechanical means is more effective than the removal of Cu(II). At low H202 concentrations, the surface is largely composed of copper metal and cuprous oxide species. Thus, increasing H202 concentration leads to higher removal rates as more easily removable Cu(I) species gain more surface coverage. However, from a certain peroxide concentration onwards, the formation of Cu(II) species is favored at the surface. Since the Cu(II) species are not as easily removed as the Cu(I) species, the removal rate decreases to a constant value associated with a fully covered surface by Cu(II) oxides/hydroxides. In addition, previous studies showed that the thickness of the oxide/hydroxide layer grows with increasing the H202 concentration. Thus, once the entire surface is covered with copper oxides/hydroxides, the removal rate remains constant.
The present results indicate that the addition of sorbate is effective at low and intermediate concentrations of H202, and decreases the copper removal rate with such compositions. At these concentrations, the copper oxides/hydroxides did not cover the entire copper surface (as seen in HRSEM micrographs obtained in 0.3 and 1 percent by volume H202 solutions, i.e., Figure 4). As a result, the exposed copper surface which was not covered by the oxides/hydroxides was available to react with sorbate anions to form Cu-O-C bonds adjacent to the copper surface. These bonds can be formed via the oxygen atoms in the sorbate carboxylic group, as indicated in previous studies for Cu- O-C bond. Thus, the overall removal rate decreased with the addition of sorbate. At high concentrations of H202 (see, Table 2), the fast formation of the copper oxides/hydroxides prevented sorbate anions to form a bond adjacent to the copper surface. Thus, the addition of sorbate did not affect the copper removal rate. However, it is known the sorbate anions can also be adsorbed on the copper oxides/hydroxides.
At high concentrations of H202, it seems that sorbate has no effect on the removal rate. Nonetheless, wettability tests performed on copper samples that were polished in slurries containing high concentration of peroxide showed an increase in the water contact angle from 51° to 69° with increase in the sorbate concentration, as discussed hereinbelow. This result may imply that sorbate adsorbs on the oxides/hydroxide surface. The Effect of Glycine on the Copper Removal Rate:
Figure 7 presents a comparative plot of three copper removal rate curves using CMP slurries containing three different H202 concentrations and containing K-sorbate at a concentration of 16 g/L, each curve plotted as a function of glycine concentration and 3 different H202 concentrations.
As can be seen in Figure 7, the copper removal rate exhibited a different behavior according to the H202 concentration in the slurry. In all 3 H202 concentrations, increasing the glycine concentration from 1 to 20 g/L, increased the cooper removal rate. However, at low and high concentrations of H202 (0.3 and 5 percent by volume, respectively), a further increase in the glycine concentration had an insignificant effect on the removal rate, even at high glycine concentrations. At intermediate concentrations of H202 (1 percent by volume), a further increase in the glycine concentration gradually increased the removal rate.
Table 3 presents the copper removal rate values obtained in slurries containing different glycine and sorbate concentration ratios and 0.3 percent by volume H202.
As previously indicated, at low concentrations of H202, the copper surface is composed of mainly copper metal with Cu(I) species as cuprous oxide. Therefore, oxidized copper ions are available at the exposed copper surface to react with either the glycine to form Cu-glycine soluble complexes, or with the sorbate to form a protective Cu-sorbate layer.
Table 3
Figure imgf000042_0001
As can be seen in Table 3, the removal rate values obtained with different concentrations of glycine and sorbate suggest that the removal rate at low H202 concentrations region is determined by relative reaction rates of Cu-glycine complexation and Cu-sorbate formation. In solutions containing concentrations above 20 g/L of glycine, any addition of glycine did not have any significant effect on the removal rate. This behavior suggests that all the available sites at the exposed copper surface were already occupied with a certain amount of either sorbate or glycine ions.
As can further be seen in Table 3, the four-fold increase in the sorbate concentration had a three-fold attenuating effect on the copper removal rate, demonstrating the rapid re-passivation effect of sorbate.
Similarly, at high H202 concentration, the copper surface is mostly covered with Cu(II) species. As mentioned hereinabove, at high concentrations of H202, the rapid formation of copper oxides/hydroxides prevents sorbate anions to bond directly to the copper surface, hence an insignificant effect of sorbate on the removal rate is observed (see, Table 2). This behavior suggests that the removal rate at high H202 concentration region is determined by relative reaction rates of Cu-glycine complexation and copper oxides/hydroxides formation. Similarly to low H202 concentration, in solutions containing concentrations above 20 g/L of glycine, any addition of glycine to the slurry had no impact on copper removal rate.
At intermediate concentrations of H202, the copper surface is composed of both Cu(I) and Cu(II) species (partially covered with Cu(I) and Cu(II) oxides/hydroxides). As opposed to low and high H202 concentrations, at intermediate concentrations of H202, increase in the concentration of glycine above 20 g/L led to a gradual increase in the copper removal rate values. This behavior suggests that the copper removal rate is determined by relative reaction rates of Cu-glycine complexation, Cu-sorbate and copper oxides/hydroxides formation. The effective area at the copper surface which was previously occupied by either sorbate or Cu(II)oxides/hydroxides (at low and high concentrations of H202, respectively) was diminished, as a portion of it was occupied by Cu(I) oxides.
As mentioned hereinabove, the Cu(I) species are more susceptible to both etching and mechanical removal. Thus, the removal rate at intermediate concentrations of H202 is higher compared to the removal rate obtained from the same glycine concentration at low or high H202 regions. EXAMPLE 3
XPS MEASUREMENTS
X-ray photoelectron spectroscopy (XPS) analysis of the copper surface was performed in order to determine the surface composition subsequent to the CMP process. The samples were polished with CMP slurries containing 64 g/L K-sorbate, 10 g/L glycine and varying concentrations of H202.
Different H2O2 Concentrations - Surface Mode:
Figures 8A-D present comparative plots of high-resolution X-ray photoelectron spectra (XPS) of Cu2p3/2 obtained from copper surface subsequent to CMP process in slurries containing 64 g L K-sorbate, 10 g L glycine and three different H202 concentrations (0.1, 1 and 10 percent by volume presented in Figures 8C, 8B and 8 A respectively) and recorded at a take-off angle of 12° relative to the copper surface, whereas etched copper sample spectrum serves as a reference (Figure 8D).
As can be seen in Figure 8, the Cu2p3/2 spectrum recorded in the etched sample is composed of a single peak located at 932.8eV and attributed to both a cuprous oxide and a metallic copper. As the H202 concentration increased from 0 to 0.1 and to 1 percent by volume, an energy shift of 0.5 eV in the Cu2p3/2 main peak was observed. A further increase in the H202 concentration to 10 percent by volume shifted the main peak back to 932.8 eV, as recorded in the etched sample.
Apart from the Cu2p3/2 main peak, two additional peaks accompanied with satellite peaks appeared in the spectrum obtained from the copper samples polished in the H202 slurries.
The presence of the satellite peaks at 940-945 eV verify the presence of Cu(II) species. In the spectra obtained from the copper samples polished in 0.1 and 1 percent by volume H202 slurries, the additional peaks are positioned at binding energies of 934.8 and 936.3 eV. These peaks can be referred to CuO and Cu(OH)2, respectively. Attributing these peaks to CuO and Cu(OH)2 can be further verified by Ols spectra. In the spectrum obtained from the copper sample polished in the 10 percent by volume H202 slurry, the additional peaks located at 933.8eV and 935. leV can be unambiguously attributed to CuO and Cu(OH)2, respectively. Further investigations of the Ols spectra provided additional information on the oxidized copper species, which coincided with the results obtained from the Cu2p3/2 spectra.
Figures 9A-C present comparative plots of high-resolution X-ray photoelectron spectra (XPS) of Ols obtained from copper surface subsequent to CMP process in slurries containing 64 g/L K-sorbate, 10 g/L glycine and different H202 concentrations (0.1, 1 and 10 percent by volume presented in Figures 9C, 9B and 9 A respectively).
As can be seen in Figures 9A-C, four peaks were observed in the spectrum obtained from the copper sample polished with the 0.1 percent by volume H202 slurry, located at 531.1, 531.7, 532.5 and 535.2 eV. The peak located at 531.1 eV was assigned to cuprous oxide while the peak positioned at 531.7 eV confirmed the presence of Cu(OH)2. The peak at 532.5 eV was assigned to C-0 species, possibly attributed to Cu-sorbate and the small peak at 535.8 eV can be attributed to adsorbed water.
As H202 concentration increased, the peak related to Cu20 shifted to lower binding energies, related to CuO. The precise intensity of the cupric oxide layer (CuO) cannot be extracted from the Ols spectrum due to a small chemical shift, compared with Cu20. However, broadening of the peak to lower binding energies associated with CuO may indicate the formation of CuO, as can be clearly seen in the Cu2p3 2 spectra.
Composition Profile of the Protective Layer:
Figures 10A-B present Cu2p3 2 (Figure 10A) and Ols (Figure 10B) X-ray photoelectron spectra (XPS) at different take-off angles taken from the surface of a copper sample subsequent to CMP process in slurry containing 64 g/L K-sorbate, 10 g/L glycine, and low H202 concentration of 0.1 percent by volume. The take-off angle of the bulk mode (69° related to the copper surface) provides maximum information from the bulk.
As can be seen in Figures 10A-B, the Cu2p3/2 spectrum recorded at a take-off angle of 69° consists of only the main peak shifted to a higher binding energy with two correlating peaks assigned to Cu20 and C-0 in the Ols spectrum. These results suggest that the energy shift in the Cu2p3 2 spectrum was attributed to an induced dipole at the copper surface. Without being bound to any particular theory, the dipole may have resulted due to the formation of a copper-oxygen association between the polished copper surface and an oxygen atom in the sorbate molecule. This association may be formed adjacent to the exposed surface areas during the CMP in 0.1 and 1 percent by volume H202 slurries.
An energy shift of the Cu2p3/2 main peak was previously observed in an earlier XPS study for copper poly-ethylene terephthalate interface, attributed to a dipole induced by the formation of Cu-O-C bond. As the take-off angle decreased, the relative intensity of Cu(OH)2 and CuO slightly increased, while the relative intensity of Cu20 decreased. This behavior indicates a higher surface concentration of Cu(OH)2 and CuO and a higher in-depth concentration of Cu20. The relative intensity of the peaks attributed to the Cu-sorbate species increased with a decrease in the take-off angle; nevertheless, these sorbate surface film species appear in all take-off angles spectra.
This observation indicates that the Cu-sorbate species are a major component in the protective film formed at the copper surface as a result of the CMP process in low concentrations of H202.
Figures 11A-B present Cu2p3/2 (Figure 11A) and Ols (Figure 11B) X-ray photoelectron spectra (XPS) at different take-off angles taken from the surface of a copper sample subsequent to CMP process in slurry containing 64 g/L K-sorbate, 10 g/L glycine, and high H202 concentration of 10 percent by volume.
As can be seen in Figure 11 A, the peak fit of the Cu2p3/2 spectra indicates that the relative intensity of Cu(OH)2 and CuO increased with a decrease in the take-off angle, indicating a higher concentration of Cu(OH)2 and CuO at the top surface level of the protective layer. In contrast, a decrease in the relative intensities of Cu20 peaks was observed as the take-off angle decreased, indicating a higher in-depth concentration of Cu20. The results obtained from the Ols spectra coincide with those obtained in the Cu2p3 2 spectra.
As can further be seen in the Ols spectrum of Figure 11B, the peak intensity at
532.5 eV, assigned to the Cu-sorbate bond increased with a decrease in the take-off angle, indicates a higher concentration of Cu-sorbate at the top surface levels of the protective layer. From the Ols spectra recorded at a take-off angle of 69°, it can be clearly seen that the peak assigned to the Cu-sorbate bond (532.5 eV) did not appear in the sample polished in the 10 percent by volume H202 slurry, but appeared in the sample polished in the 0.1 percent by volume H202 (see, Figure 10). These results indicate that the Cu-sorbate bond was not formed adjacent to the copper surface at high concentrations of H202, but rather adsorbed on the copper oxide/hydroxide layer. These results support the removal rate values, which decreased with the increase in sorbate concentration at low and intermediate H202 concentrations, while at high H202 concentration, the removal rate was not affected by sorbate concentration in the slurry.
Figure 12 presents comparative plots of extrapolated data estimated from the intensity ratio (Io/Im) between the oxide and the metal peaks obtained in XPS measurements, showing the oxide thickness developed on copper as a function of H202 content in the slurry. The oxide thickness was estimated using Strohmerier [Strohmeier, B.R., Surf. Int. Anal. 15 (1990), 51] and Seah [Seah, M. et al, Surf. Int. Anal. 1 (1979), 2] equations.
As can been in Figure 12, the copper oxide layer is mainly composed of Cu20 with a thickness of about 0.8 nm at low concentration of H202 (0.1 percent by volume), while at a high H202 concentration (10 percent by volume), the copper oxide layer is mainly composed of CuO with a thickness of about 2.8 nm. As can further be seen in Figure 12, not only the oxide composition is modified with an increase in the H202 concentration, but also its thickness. The thickness and layer composition would have a major effect on both the copper dissolution and passivation during the CMP process.
EXAMPLE 4
CONTACT ANGLE MEASUREMENTS
Static contact angles of de-ionized water at the planarized copper surface were used to indicate and quantify the change in the surface tension of a copper-containing substrate.
H202 Effect:
Figures 13A-D present contact angle measurements of a water droplet at a copper wafer surface (Figures 13A and 13B) and SEM images (Figures 13C and 13D) obtained from the copper wafer surface, subsequent to CMP process in slurries containing 64 g L K-sorbate and 10 g/L glycine with H202 concentrations of 0.1 percent by volume (Figures 13 A and 13C) and of 5 percents by volume (Figures 13B and 13D). As can be seen in Figures 13 A and 13B, the increase in the H202 concentration from 0.1 to 5 percent by volume decreased the contact angle from 68 ° to 53 °, indicating that there was a change in the surface chemistry, which altered its hydrophilicity.
Figure 14 presents comparative plots showing the XPS intensities ratio of
Cu(OH)2 and CuO, normalized to the Cu° peak in the Cu2p3/2 spectrum, as measured at an angle of 12° related to the copper surface subsequent to polishing with slurries containing 64 g/L K-sorbate, 10 g/L glycine and different H202 concentrations.
As can be seen in Figure 14, the relative intensity of Cu(OH)2 and CuO peaks increased with the H202 concentration. These results are in agreement with the HRSEM micrographs obtained earlier (Figures 4A, 4C and 4E), indicating that the increase in the H202 concentration leads to a higher coverage of the copper oxides/hydroxides at the copper surface.
As can further be seen in Figure 14, the contact angle decreased with the increase in CuO and Cu(OH)2 relative intensities. These results suggest that copper hydroxides actually induce the hydrophilicity. In general, metal oxide is less hydrophilic than the metal itself, and thus wettability may become lower with an increase in the amount of oxides at the surface. However, the wettability can be enhanced with an increase in the amount of free hydroxyl ions that can easily be combined with water. Furthermore, the copper surface obtained subsequent to the CMP process in the slurry containing 5 percent by volume H202 promotes the adhesiveness of silica particles at the surface (see, Figure 13C and 13D).
XPS results showed an increase in Cu(OH)2 intensity as the H202 concentration in the slurry increases. Both the decrease in the contact angle and the enhanced adhesiveness of the silica particles may suggest that Cu(OH)2 is positioned at the upper copper surface layer. Therefore, the adhesion of silica particles may be attributed to hydrolysis and condensation reaction at the Cu(OH)2 layer.
Sorbate Effect:
Figures 15A-D present contact angle measurements of a water droplet at a copper wafer surface (Figures 15A and 15B) and SEM images (Figures 15C and 15D) obtained from the copper wafer surface, subsequent to CMP process in slurries containing 10 g/L glycine and 0.1 percent by volume H202 with sorbate concentrations of 4 g/L (Figures 15 A and 15C) and of 64 g/L (Figures 15B and 15D).
As can be seen in Figures 15A and 15B, the increase in the sorbate concentration from 4 to 64 g/L in slurries containing low concentration of H202, increased the contact angle from 80° to 99°, indicating the presence of the fatty acid (sorbate) at the copper surface. The sorbate molecule is a 6-carbons fatty acid chain having hydrophobic alkyl tails; thus, the decrease in the wettability, as sorbate concentration increased was most probably due to the sorbate hydrophobic tail.
As can further be seen in Figures 15C and 15D, there is a substantial reduction in the silica particles' adhesiveness with the increase in the sorbate concentration.
In slurries containing high concentrations of H202, the increase in sorbate concentration led to an increase in the water contact angle, from 51° to 69° (data not shown). This observation coincides with the XPS results, implying that the Cu-sorbate species are a major component of the bulk protective film, formed at the copper surface at low H202 concentration. However, at high H202 concentration, sorbate is most probably adsorbed mainly at the copper oxide/hydroxide layer. Thus, the increase in water contact angle is less pronounced in high H202 concentration slurries once compared with data obtained from slurries containing low H202 concentration.
It is noted that while difference in the increase of the contact angle value between the low concentration of H202 and the high concentration of H202 may not be significant, the absolute values carry significance; in low H202 concentration the contact angle increased to a max value of 99 while in high H202 concentration it increased to 69 due to different layer composition (in which the layer composition consists of higher oxides/hydroxides ratio than in low H202 concentration case).
Figures 16A-D summarize the postulated processes occurring on the surface of copper when using the H202-glycine slurries containing sorbate, as concluded from the experimental results discussed hereinabove. Figures 16A-D are schematic diagrams depicting a theoretical surface of a copper sample polished by slurries having two concentrations of H202, 0.1-0.3 percent by volume (low H202, Figures 16A and 16B) and 5-10 percent by volume (high H202, Figures 16C and 16D), and two sorbate concentrations, 4g/L (low sorbate, Figures 16A and 16C) and 64 g/L (high sorbate, Figures 16B and 16D). The sample depicted in Figure 16 A can be characterized by a copper removal rate of 240 nm/min and a contact angle of 80 °; the sample depicted in Figure 16B can be characterized by a copper removal rate of 83 nm/min and a contact angle of 99 °; the sample depicted in Figure 16C can be characterized by a copper removal rate of 234 nm/min and a contact angle of 51 °; and the sample depicted in Figure 16D can be characterized by a copper removal rate of 246 nm/min and a contact angle of 69 °.
As can be seen in Figures 16A and 16B, at a low concentration of H202 a thin layer or possibly islands of copper oxide, composed mainly from Cu20, are formed at the copper surface alongside with Cu-sorbate species; however, the oxide layer does not completely cover the copper surface. Following the mechanical abrasion, the copper is exposed to the slurry and the removal rate is determined by both the glycine and the sorbate concentration in the slurry. Thus, the copper removal rate decreases with an increased sorbate concentration.
As can be seen in Figures 16C and 16D, at a high concentration of H202, a thick layer of CuO and Cu(OH)2 (with a thin under-layer or possibly a mixture of Cu20) is formed at the copper surface and covers larger portions of the copper surface. Contact angle values obtained subsequent to the CMP process with different sorbate concentrations in the slurry suggest that sorbate anions are adsorbed at the copper oxides/hydroxides surface. Following the mechanical abrasion, fresh copper is exposed to the slurry and the removal rate is determined by the relative reaction rates of both Cu- glycine complexation and copper oxides/hydroxide formation.
EXAMPLE 5
CMP OF BLANKET WAFERS CMP studies of copper blanket wafers were performed with slurries containing various concentrations of sorbate (0 to 64 g/L), peroxide (0.1 percent by volume to 10 percent by volume), and a fixed amount of glycine (10 g/L).
According to the mechanism suggested hereinabove, the surface composition and thickness are being altered according to the constituent's concentration in the slurry, consequently determining copper removal rate and wafer performance.
Figure 17 presents a 4 x 5 array of 20 optical microscope images of copper surface subsequent to CMP process with CMP slurries containing different concentration of sorbate and H202 and a fixed glycine concentration of 10 g/L, denoting the removal rate (RR in nanometer per minute, nm/min) values, wherein the concentration of H202 increases column-wise 0.1, 0.3, 1, 5 and 10 percents by volume, and the concentration of sorbate increases row- wise 0, 4, 16 and 64 g/L.
As can be seen in Figure 17, at low H202 concentrations, the surface is mainly covered with islands, or possibly a thin layer, of cuprous oxides (0.8 nm) alongside Cu- sorbate species formed on top of the copper surface. Without being bound by any particular theory, it is assumed that the oxide layer does not cover the entire surface or is not thick enough to protect the copper surface. Thus, at low concentrations of H202, the surface is characterized with a coarse roughness, indicating uncontrolled etching of copper. With the addition of sorbate, the surface exhibits a finer topography with a reduced roughness.
XPS results revealed that at low H202 concentrations, Cu-sorbate species are a major component in the protective layer. Thus, the "smeared-like" surface may be explained by the presence of organic chain molecules which lubricate the surface and cannot provide a planar surface.
At high H202 concentration, the copper surface is mainly composed of cupric oxides and hydroxides (about 4 nm). Thus, the roughness decreases as the etching of the surface occurs in a more controlled manner. With the addition of sorbate, the oxide/hydroxide layer is additionally covered with adsorbed sorbate anions and the surface exhibits an improved roughness, even with the addition of low sorbate concentrations.
Hence, combinations of high concentrations of H202 and different sorbate concentrations (marked by a red frame in Figure 17) provide smooth copper surface at a high copper removal rate. The slurry compositions which afforded the copper polishing indicated by a red frame in Figure 17 were further studied in order to optimize the effect of sorbate on copper CMP performance.
EXAMPLE 6
AFM MEASUREMENTS
AFM measurements were performed on copper surface subsequent to CMP with slurries containing 10 g L glycine, 5 percent by volume H202 and different concentrations of sorbate, in order to quantify the effect of the sorbate concentration on the roughness during the copper CMP process itself.
As known in the art, chemical passivation, chemical etching and mechanical abrasion are some of the main driving forces for the removal of the metal in the CMP process. Initially, the abrasive particles remove the protective oxide film from the protruded areas on the wafer. As a result from the abrasive action, underlying bare copper is exposed to the slurry. Throughout the alternating cycles of chemical formation and mechanical removal of a surface film, the protruded areas are exposed, while the recessed areas are still covered by the pristine oxide layer, thus being in contact with only the CMP slurry. The recessed (sunk) areas must be protected as otherwise isotropic dissolution would occur along the uneven copper structure and planarization would not be achieved.
Figures 18A-B present AFM micrographs of two copper samples subsequent to polishing with slurries having 4 g/L (Figure 18A) and 64 g/L (Figure 18B), wherein the roughness of the sample treated with 4 g/L sorbate exhibited a Z range of 37.319 nm, an
RMS (Rq) of 1.621 nm, and a mean roughness (Ra) of 1.306 nm, and the sample treated with 64 g/L sorbate exhibited a Z range of 13.533 nm, an RMS (Rq) of 1.284 nm, and a mean roughness (Ra) of 0.99 nm.
As can be seen in Figures 18A-B, increasing the sorbate concentration from 4 g/1 to 64 g/1, decreases substantially the height range (Z range) from a value of 37 nm to 13 nm in the measurement area.
The decrease in the height range with increased sorbate concentration indicates that sorbate anions are adsorbed to the native oxide layer at the recessed areas and protect it from dissolving into the slurry. Thus, while the exposed copper at the protruded areas dissolve into the slurry, better planarization is achieved.
EXAMPLE 7
CMP of Patterned Wafers
CMP studies on patterned wafers were performed with slurries containing different concentrations of sorbate and fixed concentrations of glycine (10 g/L) and H202 (5 percent by volume). Patterned wafer are used extensively by semiconductor capital equipment manufacturers to develop new wafer fabrication processes and maintain existing ones. A typical CMP characterization wafers are used to evaluate the performance of various CMP products such as slurries. Patterned test wafers differ from blanket film test wafers because they include additional process steps to create patterns on the blanket film and contain several arrays with copper/barrier features of different sizes and spacing. The die pattern can include many test structures, including lines, spaces, holes and different features. Hence, a patterned test wafer simulates the behavior of device wafers processed under similar conditions.
Wafer performance and polishing uniformity:
Figures 19A-D present optical microscope images of copper bond pads (100 x 100 μπι ) subsequent to copper and barrier CMP, using copper CMP slurries having 4 g/L sorbate (Figures 19A and 19C) or 64 g L sorbate (Figures 19B and 19D), polished by an edge die (Figures 19A and 19B) or a center die (Figures 19C and 19D), demonstrating the surface finish and polishing uniformity as a function of the sorbate concentration.
As can be seen in Figures 19A-D, at the lowest concentration of sorbate (4 g/1), copper was removed from the structures ("dished-out") and a difference in polishing between center and edge die was observed also in Figure 19C. These results could imply that the hydroxides/oxides, formed at the copper surface, were not fully covered by sorbate anions and consequently dissolved. In this perspective, a partial coverage of the wafers by the protective sorbate film might have lead to a non-uniform polishing. One can conclude that in these conditions 4 g/L sorbate in the CMP slurry is not a sufficient concentration. At high concentrations of sorbate (64 g 1), the amount of sorbate anions capable of covering a larger portion of the wafer surface is more than sufficient and therefore, the copper surface is being uniformly passivated.
Without being bound by any particular theory, it is assumed that the film produced at the copper surface is composed of hydroxides, oxides and sorbate multilayer in an overall thickness of 4-10 nm. As no monolayer is being formed, one would need a sufficient dissolved sorbate in the solution in order to establish a full coverage (and hence an efficient protection) of the copper surface.
These results indicate that the addition of sorbate to H202-glycine solutions expands the passivity region and provides an enhanced copper passivity due to the formation of both oxide and sorbate protective film at the copper surface. Furthermore, these results illustrate the essence of using inhibitors having well suited characteristics for CMP, i.e. the stability of the formed protective film in a wide variety of media and fast re-passivation kinetics, as the formed copper oxides/hydroxides cannot provide the required passivation.
Figures 20A-D present SEM images of 90 nm copper wiring, showing top view
(Figures 20A and 20B) and side view (Figures 20C and 20D), polished with CMP slurries having 4 g L sorbate (Figures 20A and 20C) or 64 g/L sorbate (Figures 20B and 20D), wherein copper corrosion sites are marked by arrows in Figure 20C.
As can be seen in Figures 20A-D, an increase in sorbate concentration leads to a decrease of copper corrosion sites along the copper lines, and at a high concentration of sorbate, copper lines are more protected while at the lower concentration copper pits are observed.
EXAMPLE 8
DISHING VALUES AS A FUNCTION OF SORBATE CONCENTRA TIONS BY HIGH RESOLUTION PROFILOMETER (HRP) MEASUREMENTS The HRP test was performed in order to evaluate sorbate performance during the over-polishing stage. Dishing values of CMP structures were evaluated subsequent to copper CMP with the selected slurries.
Figure 21 presents the dishing values of copper van der-Pauw (VDP) structures
(80 μιη x 80 μπι) obtained as a result of a CMP process performed under the conditions presented in Table 1 hereinabove, plotted as a function of sorbate concentration in the slurry.
As can be seen in Figure 21, dishing values decrease substantially with increase in sorbate concentration. The decrease in dishing values with increased sorbate concentration indicates that sorbate anions were adsorbed at the oxide layer and covered a larger portion of the surface thus provided an enhanced protection to the copper lines and structures. These results coincide with XPS and contact angle results obtained for these slurries and presented hereinabove (see, Example 4). EXAMPLE 9
COPPER SHEET RESISTANCE
Different Pattern density (PD):
Figure 22 presents a box plot of normalized copper sheet resistance values at different copper densities (PD) as a function of sorbate concentrations, wherein the sheet resistance values are normalized to a maximum value obtained in sample that was polished with a slurry having a sorbate concentration of 4 g/L.
As can be seen in Figure 22, higher sheet resistance values are obtained in high PD structures, and at both 10 % and 60 % PD, increased sorbate concentration leads to a drastic decrease in sheet resistance, indicating substantial low copper loss. Wafer non- uniformity is also markedly reduced as the sorbate concentration increases, indicating on narrow distribution of sheet resistance values across the whole wafer, most probably due to higher coverage of sorbate anions at the copper oxides/hydroxides layer. These results coincide with previous observation on a uniform polishing obtained from increased sorbate concentration slurries (see, Example 5).
Different line width:
Figure 23 presents a box plot of normalized copper sheet resistances values as a function of sorbate concentration in the slurry measured from different line widths, wherein the sheet resistance values are normalized to a maximum value obtained from sample polished with a slurry having a sorbate concentration of 4 g/L.
As can be seen in Figure 23, higher copper sheet resistances values are obtained in wider lines, and at 1 μπι and 30 μπι LW, increasing the sorbate concentration leads to a decrease in sheet resistance, indicating lower dishing.
Figure 24 summarizes the CMP process with respect to sorbate concentration in a 5 percent by volume H202-glycine based slurry as concluded from the experimental results presented hereinabove. Figures 24A-D are schematic diagrams depicting a theoretical surface of a copper sample polished by slurries having two sorbate concentrations, 4 g/L (low sorbate, Figures 24A and 24C) and 64 g/L (high sorbate, Figures 24B and 24D), in two polishing stages, the main CMP process (Figures 24A and 24B) and the over-polishing stage (Figures 24C and 24D).
The sample depicted in Figure 24A can be characterized by a height range of 37 nm; the sample depicted in Figure 24B can be characterized by a height range of 13 nm; the sample depicted in Figure 24C can be characterized by a dishing value subsequent to copper CMP process of 140 nm; and the sample depicted in Figure 24D can be characterized by a dishing value subsequent to copper CMP process of 50 nm.
Referring to Figures 24A-B, during the copper CMP process, sorbate anions are adsorbed to the copper surface at the recessed areas, assisting in achieving a better planarization, while at low sorbate concentration the recessed areas are not well protected and uncontrolled etching occurs along the copper lines. At high sorbate concentrations, the recessed areas are protected and etching is restrained, therefore, at high sorbate concentrations, lower height range and RMS roughness values are recorded (AFM measurements).
Referring to Figures 24C-D, during the over-polishing step and barrier CMP, the copper structures should be protected to avoid further etching of the planarized structures, which could result in open circuit of the device. The protective layer exposed to the slurry is composed of mainly CuO and Cu(OH)2. At low sorbate concentrations, copper oxides and hydroxides are not well protected and dissolve in the presence of glycine and acidic media, which is typically used in subsequent steps, further down in the production line of ICs. At high sorbate concentrations, copper oxides/hydroxides films are well covered with protective adsorbed sorbate anions therefore, lower dishing values are obtained.
EXAMPLE 10
POST CMP SURFACE "CLEANNESS "
In order to further demonstrate the effect of sorbate on post-CMP adhesion of abrasive particles on the surface of polished copper, as discussed hereinabove in Example 4 (see, e.g., Figures 13C and 13D), copper-containing wafer samples were subjected to CMP using an exemplary slurry composition containing 10 g/L glycine, 10 percents by volume H202 and silica abrasive particles. The copper-containing wafer samples were thereafter subjected to post-treatment with a similar composition with the exception of containing 0.4 M sorbate.
Electron micrographs were taken before and after the post-treatment with sorbate, and the results are presented in Figures 25A-D. Figures 25A-D are electron micrographs at a magnification of 8,000X (Figure 25 A), 64,000X (Figure 25B), 10,000X (Figure 25C) and 50,000X (Figure 25D) of copper-containing wafer samples subjected to CMP before post- treatment with 0.4 M sorbate (Figures 25A and 25B) and after post-treatment with 0.4 M sorbate (Figures 25C and 25D), showing significant decrease in the silica particles adhesiveness as a result of the surface passivation effect imparted to the copper surface by sorbate.
CONCLUSIVE REMARKS
Results obtained from electrochemical, contact angle, XPS and CMP studies presented hereinabove provide an insight into the mechanism involved in copper CMP conducted with H202-glycine slurry containing sorbate as an inhibitor, and allow determination of a CMP slurry composition with superior performance as compared to currently known CMP slurries. Electrochemical studies show that the addition of sorbate to H202-glycine solutions expands the passivity region by providing an enhanced passivity due to the formation of both an oxide and a sorbate protective film at the copper surface. XPS results reveal that both the surface composition and the thickness of the protective layer alter as H202 concentration increases.
Removal rates values obtained in the CMP studies on blanket wafers suggest that the composition, thickness and coverage of the oxide/hydroxide film formed at the copper surface determine the possible reactions, as well as their rate according to the concentrations of the various ions present in the slurry (sorbate and glycine). Furthermore, it was found that sorbate repel silica particles at the copper surface. The ability of sorbate to provide both a passivated and a silica-free copper surface can facilitate the next cleaning step (post-CMP step).
In addition to the ability of sorbate to provide an enhanced copper passivation during electrochemical measurements, the results presented hereinabove reveals that copper CMP slurry containing an increased sorbate concentration yields lower roughness values on copper blanket wafers. On patterned wafers, reduced copper losses were found with increased sorbate concentrations.
It has also been demonstrated that the chemical composition of copper slurry has an impact on copper dishing; its origin is not only found in polishing pad asperities. The results obtained and presented hereinabove indicate clearly that the type and concentration of the inhibitor used in the slurry has a major influence on the obtained dishing values and CMP performance, and even more so the results indicate that sorbate can be used to better optimize copper CMP process, by controlling dishing phenomenon.
Hence, according to some embodiments of the invention presented herein, an exemplary beneficial sorbate anion concentration in the CMP slurry composition is 0.4 M, the concentration of glycine is 20 grams/L, and a concentration of hydrogen peroxide is 10 percents by volume.
Although the invention has been described in conjunction with specific embodiments thereof, it is evident that many alternatives, modifications and variations will be apparent to those skilled in the art. Accordingly, it is intended to embrace all such alternatives, modifications and variations that fall within the spirit and broad scope of the appended claims.
All publications, patents and patent applications mentioned in this specification are herein incorporated in their entirety by reference into the specification, to the same extent as if each individual publication, patent or patent application was specifically and individually indicated to be incorporated herein by reference. In addition, citation or identification of any reference in this application shall not be construed as an admission that such reference is available as prior art to the present invention. To the extent that section headings are used, they should not be construed as necessarily limiting.

Claims

WHAT IS CLAIMED IS:
1. A copper CMP composition comprising hydrogen peroxide, glycine and a sorbate anion, wherein a concentration of said sorbate anion in the composition is at least 0.2 mol per liter.
2. A copper CMP composition comprising:
a peroxide in a concentration that ranges from 0.1 percent by volume to 10 percent by volume;
a complexing agent in a concentration that ranges from 1 gram per liter to 40 grams per liter; and
a sorbate anion in a concentration of at least 0.2 mol per liter.
3. The composition of claim 2, wherein a concentration of said peroxide ranges from 0.5 percent by volume to 10 percents by volume.
4. The composition of any of claims 2 and 3, wherein said peroxide is hydrogen peroxide.
5. The composition of claim 2, wherein a concentration of said complexing agent ranges from 5 grams per liter to 30 grams per liter.
6. The composition of any of claims 2-5, wherein said complexing agent is selected from the group consisting of an amino acid, a monoamine-containing compound and a diamine-containing compound.
7. The composition of any of claims 2-5, wherein said complexing agent is glycine.
8. The composition of any of claims 1-7, wherein a concentration of said sorbate anion is at least 0.4 mol per liter.
9. The composition of claim 8, wherein a concentration of said sorbate anion is at least 1.7 mol per liter.
10. The composition of any of claims 1-7, wherein a source of said sorbate anion is potassium sorbate.
11. The composition of any of claims 1-10, wherein a concentration of hydrogen peroxide ranges from 0.1 percent by volume to 10 percents by volume.
12. The composition of claim 11, wherein said concentration of hydrogen peroxide ranges from 5 percent by volume to 10 percents by volume.
13. The composition of claim 11, wherein said concentration of hydrogen peroxide ranges from 0.3 percent by volume to 1 percent by volume.
14. The composition of any of claims 1 and 7-10, wherein a concentration of glycine ranges from 5 grams per liter to 40 grams per liter.
15. The composition of claim 14, wherein said concentration of glycine ranges from 5 grams per liter to 30 grams per liter.
16. The composition of claim 14, wherein said concentration of glycine ranges from 10 grams per liter to 30 grams per liter.
17. The composition of any of claims 1 and 7, wherein the concentration of said sorbate anion ranges from 0.4 mol per liter to 9 mol per liter, a concentration of glycine is 20 grams per liter, and a concentration of hydrogen peroxide ranges from 5 percent by volume to 10 percents by volume.
18. The composition of any of claims 1-17, having a pH that ranges from 7 to 10.
19. The composition of claim 18, having a pH of 8.5.
20. The composition of any of claims 1-19, further comprising abrasive particles.
21. The composition of claim 20, wherein said abrasive particles are selected from the group consisting of silica, alumina, titania, zirconia, ceria, germanium oxide and mixtures thereof.
22. A process of chemical-mechanical planarization of a copper-containing surface, the process comprising contacting a substrate having a copper-containing surface with the copper CMP slurry composition of any of claims 1-21, thereby planarizing the copper-containing surface.
23. The process of claim 22, further comprising, subsequent to said contacting, applying a polishing pad onto said substrate.
24. The process of any of claims 22 and 23, characterized by a copper removal rate of at least 200 nanometer per minute.
25. The process of any of claims 22-24, characterized by a planarized copper-containing surface that has a copper dishing value of less than 110 nanometer.
26. A substrate having a planarized copper-containing surface obtained by the process of any of claims 22-25.
27. The substrate of claim 26, wherein the planarized copper-containing surface is characterized by a height range (Z range) value of less than 20 nm.
28. The substrate of claim 26, wherein the planarized copper-containing surface is characterized by a water-to-surface contact angle of at least 80 °.
29. The substrate of any of claims 26-28, forming a part of an article that comprises a copper-containing element.
30. The substrate of claim 29, wherein said article is selected from the group consisting of an integrated circuit, a device wafer, a 3D integration device and an advanced interconnect device.
31. An article comprising a wafer-based device having at least one copper- containing element, said copper-containing element being produced by the process of any of claims 22-25.
32. The article of claim 31, being characterized by a copper dishing value of less than 110 nanometers and/or a height range (Z range) value of less than 20 nm.
PCT/IL2011/000019 2010-01-07 2011-01-06 Sorbate-containing compositions for use in copper chemical mechanical planarization WO2011083475A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US28224610P 2010-01-07 2010-01-07
US61/282,246 2010-01-07

Publications (1)

Publication Number Publication Date
WO2011083475A1 true WO2011083475A1 (en) 2011-07-14

Family

ID=43929027

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IL2011/000019 WO2011083475A1 (en) 2010-01-07 2011-01-06 Sorbate-containing compositions for use in copper chemical mechanical planarization

Country Status (1)

Country Link
WO (1) WO2011083475A1 (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5770095A (en) 1994-07-12 1998-06-23 Kabushiki Kaisha Toshiba Polishing agent and polishing method using the same
US6569350B2 (en) 1996-12-09 2003-05-27 Cabot Microelectronics Corporation Chemical mechanical polishing slurry useful for copper substrates
US20070163677A1 (en) 2003-04-10 2007-07-19 Yair Ein-Eli Copper cmp slurry composition

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5770095A (en) 1994-07-12 1998-06-23 Kabushiki Kaisha Toshiba Polishing agent and polishing method using the same
US6569350B2 (en) 1996-12-09 2003-05-27 Cabot Microelectronics Corporation Chemical mechanical polishing slurry useful for copper substrates
US20070163677A1 (en) 2003-04-10 2007-07-19 Yair Ein-Eli Copper cmp slurry composition

Non-Patent Citations (15)

* Cited by examiner, † Cited by third party
Title
ABELEV, E. ET AL., ELECTROCHEM. ACTA, vol. 52, 2007, pages 5150
ABELEV, E. ET AL., ELECTROCHIM. ACTA, vol. 52, 2007, pages 1975
ABELEV, E. ET AL., ELECTROCHIM. ACTA, vol. 52, 2007, pages 5150
ABELEV, E., ELECTROCHIM. ACTA, vol. 52, 2007, pages 1975
E. ABELEV ET AL., LANGMUIR, vol. 23, 2007, pages 11281
E. ABELEV ET AL.: "Potassium sorbate solutions as copper chemical mechanical planarization (CMP) based slurries", ELECTROCHIMICA ACTA, vol. 52, 2007, pages 5150 - 5158, XP002637207, DOI: 10.1016/j.electacta.2007.02.010 *
EIN-ELI, Y. ET AL., ELECTROCHIM. ACTA, vol. 49, 2004, pages 1499
EIN-ELI, Y. ET AL., ELECTROCHIM. ACTA, vol. 52, 2007, pages 1825
EIN-ELI, Y., J. ELECTROCHEM. SOC., vol. 151, 2004, pages 236
ELECTROCHIMICA ACTA, vol. 52, 2007, pages 6353 - 6358
SEAH, M. ET AL., SURF INT. ANAL., vol. 1, 1979, pages 2
STROHMEIER, B.R., SURF INT. ANAL., vol. 15, 1990, pages 51
THUS; PRASAD; Y.N. ET AL., ELECTROCHIMICA ACTA, vol. 52, 2007, pages 6353 - 6358
Y. EIN-ELI ET AL.: "Reviews on copper chemical-mechanical polishing (CMP) and post-CMP cleaning in ultra-large system integrated (ULSI)- An electrochemical perspective", ELECTROCHIMICA ACTA, vol. 52, 2007, pages 1825 - 1838, XP002637208, DOI: 10.1016/j.electacta.2006.07.039 *
Y. EIN-ELI, ELECTROCHEM. SOLID-STATE LETT., vol. 9, no. 5, 2006

Similar Documents

Publication Publication Date Title
JP5539934B2 (en) Chemical mechanical polishing slurry useful for copper substrate
EP2242091B1 (en) Polishing solution for metal and polishing method
TWI282360B (en) Polishing composition and polishing method thereof
WO2000000567A1 (en) Chemical mechanical polishing slurry useful for copper/tantalum substrate
WO2000000561A1 (en) Chemical mechanical polishing slurry useful for copper/tantalum substrates
JP2002519471A5 (en)
JP2005117046A (en) Abrasive composition and its use
CN109382756B (en) Chemical mechanical polishing method for tungsten
US20070293048A1 (en) Polishing slurry
CN110616044B (en) Tungsten chemical mechanical polishing composition
JP7103794B2 (en) Chemical mechanical polishing method for tungsten
EP2209864A1 (en) Polishing composition comprising a n-substituted imidazole and a method for polishing a copper film
TWI332017B (en) Abrasive-free polishing slurry and cmp process
WO2008151918A1 (en) A process for polishing patterned and unstructured surfaces of materials and an aqueous polishing agent to be used in the said process
JP6936315B2 (en) Chemical mechanical polishing method for tungsten
JP6670715B2 (en) Polishing composition used for polishing a polishing object having a layer containing metal
JP7391595B2 (en) Chemical mechanical polishing composition and method for tungsten
JP5813921B2 (en) Chemical mechanical polishing aqueous dispersion and chemical mechanical polishing method
WO2011083475A1 (en) Sorbate-containing compositions for use in copper chemical mechanical planarization
US10815392B2 (en) Chemical mechanical polishing method for tungsten
JP2003286477A (en) Polishing composition and polishing method
Jang et al. Effect of secondary inhibitors on material removal rate and nano-roughness of Cu chemical mechanical planarization
KR101203599B1 (en) Method for chemical mechanical planarization of a copper-containing substrate
JP2005056879A (en) Solution and method for polishing cooper-based metal
JP2004123931A (en) Polishing solution and polishing method

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 11703932

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 11703932

Country of ref document: EP

Kind code of ref document: A1