WO2019013761A1

WO2019013761A1 - Aqueous composition for depositing a cobalt deposit and method for electrolytically depositing such a deposit

Info

Publication number: WO2019013761A1
Application number: PCT/US2017/041475
Authority: WO
Inventors: Fred WAFULA; James Adolf; Jun Wu; Josef Gaida; Dirk Rohde; Grigory VAZHENIN
Original assignee: Atotech Deutschland Gmbh
Priority date: 2017-07-11
Filing date: 2017-07-11
Publication date: 2019-01-17
Also published as: TW201908533A

Abstract

The present invention refers to an aqueous composition for depositing a cobalt deposit, the composition comprising (a) cobalt (II) ions, (b) at least one first compound comprising an acetylenic moiety, (c) at least one polymer comprising a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen, and (d) optionally a buffering agent, and a method for electrolytically depositing a cobalt deposit onto a substrate utilizing the composition above.

Description

Aqueous composition for depositing a cobalt deposit and method for electrolytically depositing such a deposit

Field of the Invention

The present invention relates to an aqueous composition for depositing a cobalt deposit and a method for electrolytically depositing such a cobalt deposit onto a substrate, in particular to a method for electrolytically depositing such a cobalt deposit into a plurality of vias and/or trenches.

Background of the Invention

Trenches and vias are typical features on semiconductor substrates of modern electronic devices. Since copper is an excellent conducting material with basically low resistivity and high reliability, for decades and still until today, it is deposited into such features in order to fill respective vias and trenches and to form a pattern of conductive lines and interconnects. However, copper as filling metal also bears some drawbacks.

It is well known that copper has a high migration and diffusion tendency. Depending on the substrate (for example silicon oxide containing substrates), copper easily diffuses into the substrate thereby causing voids, an interrupted electrical flow and even electrical shorts. In order to address such critical problems, typically a barrier layer is applied on the surface of the substrate to block or at least significantly suppress said migration and diffusion. In many applications, barrier layers contain cobalt.

The application of such barrier layers becomes more and more demanding because during recent years the size of said features is getting smaller and smaller, demanding also thinner und more uniform barrier layers. At present, a common feature size is in the range from 50 nm to even less than 10 nm. Furthermore, in order to deposit copper electrolytically onto a barrier layer, a conductive and thin seed layer needs to be additionally deposited onto the barrier layer.

But besides such a layer arrangement, void-free filling of such small features with copper becomes increasingly demanding and difficult. In addition to decreasing feature sizes and the application of thin barrier and seed layers, feature geometry calls for additional attention because of feature's high aspect ratios. Features with an opening dimension of for example 10 nm and a feature depth of for example 100 nm exhibit an aspect ratio of 10: 1. This and similar aspect ratios become more important today and it is expected that aspect ratios will further increase in the near future. Features with such high aspect ratios require sophisticated filling methods and specifically designed copper deposition baths to avoid incomplete feature filling, causing defects widely known as voids.

Another drawback of a decreasing feature size is the resulting copper resistivity, which exponentially increases, in particular with feature dimensions below 10 nm. Although resistivity is a metal intrinsic characteristic, overall resistivity of a copper filled feature is largely affected by the metallic and non-metallic materials surrounding the copper and the way how electrons interact with these materials. For example, the flow of electrons is largely affected by coming across interfaces such as between (i) copper and embedded impurities in the copper, (ii) grain boundaries, (ii) copper and other layers such as seed and barrier layers. Recent investigations have revealed that other metals, such as cobalt, are less susceptible to such an exponential resistivity increase and it has been therefore suggested to replace copper by cobalt. This appears reasonable also for another reason. Using metals that can serve dual or even multiple purposes (as filling metal as well as barrier and/or seed layer) may allow more volume for the conducting metal and thus a lower overall resistivity. Since cobalt often serves as metal in barrier layers for copper deposits, cobalt filled features do not necessarily require an additional barrier layer. Furthermore, a cobalt layer may also serve as conductive seed layer. As a result, a very homogeneous cobalt deposit can be achieved with cobalt as copper replacement.

US 2009/0188805 A1 relates to using electrodeposition to fill recessed surface features of a substrate with metals and alloys in a substantially void free manner and discloses a cobalt deposition bath for void-free cobalt filling comprising 2-mercapto-5-benzimidaz- olesulfonic acid (MBIS) as filling additive.

CA 1086679 A relates to a process and composition for the preparation of an electro- deposit which contains cobalt. A composition may comprise an unsaturated cyclosulfone in combination with propargyl alcohol.

US 6,923,897 B1 relates to a cold rolled strip which is provided with a cobalt or a cobalt alloy layer by an electrolytic method. A bath may comprise butynediol and saccharine.

WO 2017/004424 A1 relates to electrolytic deposition chemistry and a method for depositing cobalt and cobalt alloys; and more specifically to additives and overall compositions for use in an electrolytic plating solution and a method for cobalt-based metallization of interconnect features in semiconductor substrates. A composition may comprise propargyl alcohol in combination with bis-(sodium sulfopropyl) disulphide (SPS). Today's feature sizes and aspect ratios also demand specifically designed cobalt containing compositions to obtain void free cobalt filled features, in particular organic additives supporting a good bottom-up filling of the features. Organic additives comprising an acet- ylenic moiety, such as propargyl alcohol, in some cases show an acceptable bottom-up filling performance, in particular for features with moderate aspect ratios.

An excellent bottom-up filling performance is even more required the higher the aspect ratios of features are. If the total concentration of said organic additives is too low, the bottom-up filling performance is insufficient too and voids are often observed. As a result, in many cases the total concentration of organic additives is increased in order to adequately fill features with high aspect ratios. However, if the total concentration exceeds a certain limit, inacceptable deposition defects (e.g. skip plating) frequently occur on the substrate's surface. In such a case the coverage with cobalt on the substrate's surface is insufficient. Ideally the entire surface of a substrate is completely and homogeneously covered with cobalt while all features on the substrate are void free filled with cobalt.

Objective of the present Invention

It is an objective of the present invention to provide an aqueous composition for depositing a cobalt deposit, which overcomes the above mentioned problems. It is in particular the objective, to prevent said undesired skip plating on a substrate's surface despite a comparatively high total concentration of organic additives.

It is furthermore an objective that the composition provides excellent bottom-up filling performance for cobalt in features with very small opening dimensions, preferably with comparatively high aspect ratios, and providing cobalt deposits easy to handle in subsequent processing steps.

It is an additional objective of the present invention to provide also a method for electrolyt- ically depositing a cobalt deposit. It is in particular the objective to provide a method for filling features with very small opening dimensions and preferably high aspect ratios. Furthermore an excellent coverage (i.e. no skip plating) should be obtained.

Summary of the Invention

These objectives are solved by an aqueous composition for depositing a cobalt deposit, the composition comprising

(a) cobalt (II) ions,

(b) at least one first compound comprising an acetylenic moiety, (c) at least one polymer comprising a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen, and

(d) optionally a buffering agent.

Furthermore, the additional objective is solved by a method for electrolytically depositing a cobalt deposit onto a substrate, the method comprising the steps

(A) providing the substrate,

(B) providing an aqueous composition according to the present invention (as described above and below in the present text),

(C) contacting the substrate with the aqueous composition and supplying an electrical current such that cobalt is deposited onto the substrate to obtain the cobalt deposit.

Brief description of the figure

Figures 1a and 1 b show images of cross-sections of trenches obtained from filling experiments after 20 and 30 seconds, respectively, with composition C1 ( "C" means Comparative Example).

Figures 1 c and 1 d show images of cross-sections of trenches obtained from filling experiments after 10 and 20 seconds, respectively, with composition C2.

Figure 2 shows an image of a cross-section of trenches obtained from filling experiments after 30 seconds with composition E2 ("E" means according to the invention).

Figure 3 shows an image of a cross-section of trenches obtained from filling experiments after 30 seconds with composition E4.

Figures 4a and 4b show images of cross-sections of trenches obtained from filling experiments (each after 15 seconds) with compositions E5 and E6, respectively.

Figure 5 shows an image of a cross-section of trenches obtained from filling experiments after 25 seconds with composition E10.

Figures 6a and 6b show images of cross-sections of trenches obtained from filling experiments (each after 15 seconds) with compositions E13 and E14, respectively.

Figure 7 shows an image of a cross-section of trenches obtained from filling experiments after 15 seconds with composition E16.

Figure 8 shows an image of a cross-section of a substrate with a plurality of trenches obtained after approximately 2 minutes of cobalt deposition with composition C1. In Fig. 8 to 11 "A" denotes the substrate with its trenches. "B" denotes the cobalt deposit in the trenches as well as on the substrate. Line "C" indicates the individual thickness of the cobalt deposit "B" over the substrate.

Figure 9 shows an image of a cross-section of a substrate with a plurality of trenches obtained as for Fig. 8 but with composition E5.

Figure 10 shows an image of a cross-section of a substrate with a plurality of trenches obtained as for Fig. 8 but with composition E9.

Figure 11 shows an image of a cross-section of a substrate with a plurality of trenches obtained as for Fig. 8 but with composition E13.

Detailed Description of the Invention

The composition of the present invention is an aqueous composition, which means that water is the primary component. Thus, more than 50 wt.-% of the composition is water, based on the total weight of the aqueous composition, preferably at least 70 wt.-%, even more preferably at least 90 wt.-%, most preferably 95 wt.-% or more. It is preferred that the aqueous composition is substantially free of organic solvents; more preferably does not contain organic solvents at all. Furthermore, the composition is preferably a homogeneous aqueous solution and thus preferably does not contain any particles.

The composition is for depositing a cobalt deposit, preferably a sulfur-free cobalt deposit.

Preferred is a composition of the present invention, wherein the composition is acidic, preferably has a pH in the range from 0.5 to 6.8, more preferably in the range from 1 to 6, even more preferably in the range from 1.5 to 5, most preferably in the range from 2.5 to 4.6, even most preferably in the range from 3.5 to 4.6. A basic pH is undesired because cobalt hydroxide precipitation is typically observed at a basic pH, which usually results in an undesired increased surface roughness of the cobalt deposit.

The aqueous composition of the present invention comprises, besides (a) cobalt (II) ions, a combination of organic additives, namely (b) at least one first compound comprising an acetylenic moiety and (c) at least one polymer comprising a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen. Own experiments have shown (see examples below) that the presence of the polymer prevents or at least significantly reduces said undesired skip plating on a substrate's surface despite a comparatively high total concentration of the first compound, preferably at a total concentration of 50 mg/L or more of said first compound. Furthermore, a wide range of said polymers additionally improves the bottom-up filling performance, compared to other common additives, polymers not having a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen, and polymers having a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen but of comparatively low molecular weight. In addition a very uniform cobalt deposit is obtained with said at least one polymer in addition to said at least one first compound compared to a composition not comprising said at least one polymer.

Preferred is a composition of the present invention, wherein the total concentration of the cobalt (II) ions in the composition is in the range from 0.5 g/L to 50 g/L, based on the total volume of the composition, preferably in the range from 0.7 g/L to 25 g/L, more preferably in the range from 0.9 g/L to 15 g/L, even more preferably in the range from 1.2 g/L to 1 1 g/L, most preferably in the range from 1.4 g/L to 7 g/L. A concentration below 0.5 g/L often results in an incomplete cobalt deposit and surface defects. If the concentration significantly exceeds 50 g/L undesired precipitation was observed in the composition, also frequently leading to an undesired increased surface roughness of the cobalt deposit.

The cobalt source of said cobalt (II) ions is preferably at least one cobalt salt, more preferably at least one inorganic cobalt salt and/or at least one organic cobalt salt. Preferred inorganic cobalt salts are selected from the group consisting of cobalt nitrate, cobalt sulfate and cobalt halides. Preferred cobalt halides are selected from the group consisting of cobalt fluoride, cobalt chloride and cobalt bromide. A preferred organic cobalt salt is cobalt acetate. The most preferred at least one cobalt salt is cobalt sulfate, preferably cobalt sulfate heptahydrate.

By means of the composition of the present invention a cobalt deposit is preferably obtained which primarily contains cobalt. This means that a composition of the present invention is preferred, wherein said cobalt (II) ions are the major metal ion species for metal deposition. Preferred is a composition of the present invention, wherein the total amount of said cobalt (II) ions in the composition represents 80 wt.-% to 100 wt.-% of all transition metal cations in the composition, based on the total weight of all transition metal cations in the composition, preferably at least 90 wt.-%, more preferably at least 95 wt.-%, even more preferably at least 98 wt.-%, most preferably at least 99 wt.-%.

In most of all cases the composition of the present invention is most preferably substantially free of (preferably does not contain) alloying metal cations. Thus, by means of the composition of the present invention most preferably a pure cobalt deposit is obtained.

In the context of the present invention, the term "substantially free" of a subject-matter (e.g. a compound, a material, etc.) denotes that said subject-matter is not present at all or is present only in (to) a very little and undisturbing amount (extent) without affecting the intended purpose of the invention. For example, such a subject-matter might be added or utilized unintentionally, e.g. as unavoidable impurity. "Substantially free" preferably denotes 0 (zero) ppm to 50 ppm, based on the total weight of the composition of the present invention, if defined for said composition, or based on the total weight of the cobalt deposit obtained in the method of the present invention, if defined for said deposit; preferably 0 ppm to 25 ppm, more preferably 0 ppm to 10 ppm, even more preferably 0 ppm to 5 ppm, most preferably 0 ppm to 1 ppm.

A composition according to the present invention is preferred, wherein the composition is substantially free of (preferably does not contain) nickel ions, preferably is substantially free of (preferably does not contain) nickel ions, iron ions, and copper ions, more preferably is substantially free of (preferably does not contain) nickel ions, iron ions, copper ions, aluminium ions, lead ions, and tin ions.

Preferably, the composition of the present invention is substantially free of (preferably does not contain) compounds comprising divalent sulfur and/or compounds comprising a mer- capto group, more preferably MBIS.

Furthermore, the composition is preferably substantially free of sulfur containing compounds with a sulfur atom having an oxidation number below +5, preferably below +6. Most preferably the composition does not comprise such sulfur containing compounds. This means that the composition is substantially free of (preferably does not contain) sulfur containing compounds widely used as brighteners in other metal deposition baths such as nickel plating baths. However, this does not exclude the presence of sulfate ions in the composition of the present invention because a sulfate ion contains a sulfur atom having an oxidation number of +6 (but not below +6). If the composition contains sulfur containing compounds with a sulfur atom having an oxidation number below +5 or lower in many cases sulfur is incorporated into the cobalt deposit. However, this is undesired because sulfur negatively affects resistivity in the cobalt deposit. Similar observations have been made for other alloying metals or elements. Furthermore, cobalt deposits comprising significant amounts of sulfur negatively affect further processing steps carried out after cobalt deposition, e.g. chemical mechanical polishing (CMP).

Very preferred is an aqueous composition according to the present invention with the proviso that, if the aqueous composition contains sulfur containing compounds, the only sulfur containing compounds are sulfate ions (SCU^2-)- Preferably, the composition of the present invention is for electrolytic deposition in order to obtain the cobalt deposit. Thus, the composition is preferably not for electroless cobalt deposition. Therefore, a composition of the present invention is preferred, wherein the composition does not contain effective amounts of reducing agents being capable of reducing the cobalt (II) ions to metallic cobalt. This may include that tiny amounts of such reducing agents are present in the composition. Preferably these tiny amounts are either not detectable by means of standard analytical tools or at least not detrimental to the intended utilization as an electrolyte. Preferred is a composition of the present invention, wherein the composition comprises a reducing agent being capable of reducing the cobalt (II) ions to metallic cobalt in a total concentration of 0 mg/L to 200 mg/L, based on the total volume of the composition, preferably the composition is substantially free of reducing agents being capable of reducing the cobalt (II) ions to metallic cobalt. Most preferred the composition of the present invention does not contain a reducing agent being capable of reducing the cobalt (II) ions to metallic cobalt.

Preferably, the composition is substantially free of (preferably does not contain) reducing agents being capable of reducing the cobalt (II) ions to metallic cobalt containing phosphorous, even more preferably the composition is substantially free of (preferably does not contain) phosphorous containing compounds.

In the most cases an aqueous composition according to the present invention is preferred, wherein the composition is substantially free of (preferably does not contain) an unsaturated cyclosulfone compound, preferably is substantially free of (preferably does not contain) unsaturated cyclic compounds comprising a sulfur atom.

The aqueous composition of the present invention comprises (b) at last one (preferably one) first compound comprising an acetylenic moiety.

In the context of the present invention, the term "at least one" generally denotes (and is exchangeable with) "one, two, three or more than three".

Without wishing to be bound by theory, it is assumed that the first compound acts as suppressor if the composition is utilized in the method of the present invention, such that cobalt deposition is inhibited on areas where a reduced deposition rate is desired, e.g. at the side walls inside a via or trench.

Preferred is a composition of the present invention, wherein said at least one first compound is a monomer, i.e. is not a polymer. Thus, said first compound typically has a clearly defined molecular weight. Preferably, said at least one first compound has a molecular weight in the range from 50 g/mol to 1000 g/mol, preferably in the range from 60 g/mol to 500 g/mol, most preferably in the range from 70 g/mol to 250 g/mol.

Preferred is a composition of the present invention, wherein the total number of carbon atoms in each of the at least one first compound is 30 or less than 30, preferably 25 or less than 25, more preferably 20 or less than 20, even more preferably 13 or less than 13, most preferably 6 or less than 6.

Preferred is a composition of the present invention, wherein the at least one first compound consists of carbon atoms, hydrogen atoms, and oxygen atoms.

Preferred is a composition of the present invention, wherein the at least one first compound does not contain sulfur atoms, preferably does not contain sulfur atoms and nitrogen atoms.

A composition of the present invention is preferred, wherein the at least one first compound contains at least one oxygen atom. In some cases it is preferred that the at least one oxygen atom is an ether oxygen atom. In other cases it is preferred that the at least one oxygen atom is an oxygen atom of a hydroxyl group, preferably of a terminal hydroxyl group. In the context of the present invention, the term "terminal hydroxyl group" denotes a hydroxyl group covalently connected to a -CH2- moiety.

Most preferred is a composition of the present invention, wherein the at least one first compound contains at least two oxygen atoms, wherein one oxygen atom is an ether oxygen atom and the other one is an oxygen atom of a hydroxyl group, preferably at least two oxygen atoms, wherein one oxygen atom is an ether oxygen atom and the other one is an oxygen atom of a terminal hydroxyl group.

The first compound comprises an acetylenic moiety. Preferred is a composition of the present invention, wherein the acetylenic moiety in the at least one first compound is a terminal acetylenic moiety. In the context of the present invention, the term "terminal acetylenic moiety" denotes that one of the two carbon atoms, which are connected with each other via a triple bond, is connected to a hydrogen atom. Throughout the present invention such compositions have shown best results.

Preferred is a composition of the present invention, wherein the at least one first compound is linear or branched, preferably linear.

Preferred is a composition of the present invention, wherein the at least one first compound is represented by a compound of formula (I) R¹— -=-= R²

(I),

wherein

R denotes hydrogen, C1-C14 hydroxy alkyl, or oxygen interrupted C2-C14 hydroxy alkyl,

R² denotes C1-C14 hydroxy alkyl or oxygen interrupted C2-C14 hydroxy alkyl.

"Hydroxy alkyl" as mentioned above denotes an alkyl moiety comprising one to fourteen, respectively two to fourteen, carbon atoms and at least one (preferably one) hydroxyl group. Oxygen interrupted" denotes that a respective alkyl comprises one or more than one oxygen ether atom. An oxygen interrupted C2 hydroxy alkyl is for example a -CH2- O-CH2-OH moiety. "Interrupted" means that a sequence of carbon atoms is interrupted in R and R², respectively. "C1-C14" denotes the total number of carbon atoms.

Preferably, in R and R² said C1-C14 hydroxy alkyl and oxygen interrupted C2-C14 hydroxy alkyl are independently linear or branched, preferably linear. The term "independently" denotes (e.g.) that (i) R in a compound of formula (I) is or is not individual from R² in the same compound, and (ii) a R in a first compound of formula (I) is or is not individual from another R in a second compound of formula (I) (same applies to R²).

Preferred is a composition of the present invention, wherein the oxygen interrupted C2- C14 hydroxy alkyl in R comprises 2 to 10 carbon atoms, preferably 2 to 7 carbon atoms, more preferably 2 to 5 carbon atoms, most preferably 2 to 3 carbon atoms.

Preferred is a composition of the present invention, wherein the oxygen interrupted C2- C14 hydroxy alkyl in R² comprises 2 to 10 carbon atoms, preferably 2 to 7 carbon atoms, more preferably 2 to 5 carbon atoms, most preferably 2 to 3 carbon atoms.

For preferred C1-C14 hydroxy alkyl in R and R² see below.

Preferred is a composition of the present invention, wherein in Formula (I)

R denotes hydrogen, C1-C14 hydroxy alkyl, or -(CH₂)_n-0-R³,

R² denotes C1-C14 hydroxy alkyl or -(CH₂)_n-0-R³,

wherein independently in R and R²

n is in the range from 1 to 7 and R³ denotes C1-C7 hydroxy alkyl. For the meaning of "independently" and "hydroxy alkyl" see the text above. Preferred is a composition of the present invention, wherein the hydroxy alkyl in R is C1- C10 hydroxy alkyl, preferably C1-C7 hydroxy alkyl, more preferably C1-C5 hydroxy alkyl, most preferably C1-C2 hydroxy alkyl.

Preferred is a composition of the present invention, wherein the hydroxy alkyl in R² is C1- C10 hydroxy alkyl, preferably C1-C7 hydroxy alkyl, more preferably C1-C5 hydroxy alkyl, most preferably C1-C2 hydroxy alkyl.

Preferred is a composition of the present invention, wherein the hydroxy alkyl in R³ is C1- C5 hydroxy alkyl, preferably C1-C2 hydroxy alkyl.

Preferred is a composition of the present invention, wherein independently in R and R² n is in the range from 1 to 5, preferably in the range from 1 to 3, more preferably is 1 or 2, most preferably is 1.

Preferred is a composition of the present invention, wherein the hydroxyl group in the hydroxy alkyl of R¹, R² and R³ is a terminal hydroxyl group. For the term "terminal hydroxyl group" see the text above.

In many cases preferred is a composition of the present invention, wherein R is hydrogen. In such a case the acetylenic moiety is a terminal acetylenic moiety. Preferably, the at least on first compound comprises a propargyl moiety.

Most preferred is a composition of the present invention, wherein the at least one first compound is selected from the group consisting of

Preferred is a composition of the present invention, wherein the total concentration of the at least one first compound in the composition is in the range from 5 mg/L to 250 mg/L, based on the total volume of the composition, preferably in the range from 7 mg/L to 150 mg/L, more preferably in the range from 9 mg/L to 90 mg/L, even more preferably in the range from 12 mg/L to 70 mg/L. If the concentration is significantly below 5 mg/L typically no suppressing effect is observed if a respective composition is utilized in a cobalt deposition method to fill features, usually resulting in a plurality of voids. If the concentration significantly exceeds 250 mg/L usually heavy and uncontrollable skip plating occurs in a cobalt deposition method.

The aqueous composition of the present invention comprises (c) at least one (preferably one) polymer comprising a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen. Preferably, said at least one polymer does not contain a primary amine nitrogen. A polymer typically consists of monomeric building units.

Preferred is a composition of the present invention, wherein more than 50% of all monomeric building units in the at least one polymer comprise at least one (preferably one) of said carboxamide moiety with a secondary or tertiary amine nitrogen, preferably more than 70%, even more preferably more than 85%, most preferably more than 95%. The carboxamide moieties in the monomeric building units are identical or different (for example, one species of monomeric building units comprises a carboxamide moiety with a secondary amine nitrogen and another species comprises a carboxamide moiety with a tertiary amine nitrogen). Most preferred is a composition of the present invention, wherein the at least one polymer is a homopolymer. This means that every monomeric building unit comprises identically at least one (preferably one) of said carboxamide moiety with a secondary or tertiary amine nitrogen.

Said polymers positively affect the skip plating performance on a substrate's surface despite a comparatively high total concentration of the at least one first compound. This positive effect occurs over a broad range of molecular weights (see examples in the text below, section 2.1). Thus, preferred is a composition of the present invention, wherein each of the at least one polymer has a weight average molecular weight Mw in the range from 1 200 g/mol to 50 000 000 g/mol, preferably in the range from 2 000 g/mol to 20 000 000 g/mol, more preferably in the range from 3 000 g/mol to 8 000 000 g/mol, even more preferably in the range from 4 000 g/mol to 6 000 000 g/mol, most preferably in the range from 5 000 g/mol to 5 000 000 g/mol. Furthermore, said polymers lead to a more uniform cobalt deposit over the entire substrate (see examples below, section 2.3). Said polymers but with a comparatively high molecular weight additionally improve the bottom-up filling performance in respective filling experiments (see examples below, section 2.2). Although the molecular weight is not in particular limited, it appears that this effect is primarily restricted by a lower molecular weight limit. Thus, preferred is, either for at least one or preferably for all of said at least one polymer, a weight average molecular weight Mw of at least 100 000 g/mol. Preferred is a composition of the present invention, wherein either at least one or preferably all of said at least one polymer have a weight average molecular weight Mw in the range from 100 000 g/mol to 50 000 000 g/mol, preferably in the range from 200 000 g/mol to 10 000 000 g/mol, most preferably in the range from 500 000 g/mol to 5 000 000 g/mol. Utilizing such polymers usually results in an improved skip plating performance as well as (i.e. additionally) an improved bottom-up filling performance (see examples below). Therefore, such polymers are most preferred in the composition of the present invention.

More preferred is a composition of the present invention, wherein the carboxamide moieties are independently represented by Formula (II)

(II),

wherein

R⁴ denotes a C1-C3 alkyl, an alkylene moiety connected via R⁵ to the amine nitrogen atom of the carboxamide moiety in formula (II),

R⁵ denotes hydrogen, C1-C3 alkyl, a carbon atom of the backbone chain of the polymer, or an alkylene moiety connected via R⁴ to the carbonyl carbon atom of the carboxamide moiety in formula (II), and

R⁶ denotes a carbon atom of the backbone chain of the polymer.

The moiety represented by formula (II) comprises an amine nitrogen atom, which is connected to R⁵ and R⁶, as well as a carbonyl carbon atom, which is connected to R⁴.

The term "a carbon atom of the backbone chain of the polymer" includes that the amine nitrogen atom of the carboxamide moiety in Formula (II) is either included in the backbone chain of the polymer (e.g. such as in PEOX or PMOX, see text below) or is included in the side chain of the polymer (e.g. such as in PNVA or PMVA, see text below). In R⁴ and R⁵, respectively, the term "an alkylene moiety connected via [... ]" denotes a bridging alkylene moiety leading to an intramolecular ring structure including the amine nitrogen atom and the carbonyl carbon atom (e.g. such as in PVP or PVCL, see text below). Preferred alkylene moieties comprise 3 to 5 carbon atoms (i.e. a propylene bridge, a butylene bridge or a pentylene bridge), preferably 3 carbon atoms.

Preferably, in R⁴ and R⁵ said C1-C3 alkyl is individually selected from the group consisting of methyl, ethyl, n-propy, and iso-propyl, preferably individually selected from the group consisting of methyl and ethyl.

Preferred is a composition of the present invention, wherein the at least one polymer does not contain a sulfur atom.

More preferred is a composition of the present invention, wherein each of the at least one polymer consists of carbon, hydrogen, nitrogen, and oxygen atoms.

Preferred is a composition of the present invention, wherein the at least one polymer does not contain a hydroxyl group.

Preferred is a composition of the present invention, wherein the at least one polymer does not contain an acetylenic moiety.

Preferred is a composition of the present invention, wherein the at least one polymer does not contain an ether oxygen atom.

Most preferred is a composition of the present invention, wherein the at least one polymer is selected from the group consisting of

Label Chemical name Structural formula Common abbreviation

(Ila) Polyvinylpyrrolidone PVP

n

(lib) Polyvinyl caprolactam PVCL

n (lie) Poly(2-methyl-2-oxazoline) PMOX

(lid) Poly(2-ethyl-2-oxazoline) PEOX

n

(lie) Poly(N-vinyl-acetamide) PNVA

(I if) Poly(N-methyl-N-vinyl-acetamide) PMVA

Most preferred is a composition, wherein the at least one polymer is selected from the group consisting of PVP, PEOX, and PNVA.

Preferred is a composition of the present invention, wherein the total concentration of the at least one polymer in the composition is in the range from 60 mg/L to 1000 mg/L, based on the total volume of the composition, preferably in the range from 80 mg/L to 800 mg/L, more preferably in the range from 100 mg/L to 600 mg/L, even more preferably in the range from 150 mg/L to 500 mg/L, most preferably in the range from 200 mg/L to 400 mg/L. If the concentration is significantly below 60 mg/L usually undesired skip plating in the presence of comparatively high concentrations of the first compound is no longer adequately suppressed. If the concentration significantly exceeds 1000 mg/L the deposition process is undesirably disturbed.

The composition of the present invention optionally contains a buffering agent for pH stabilization, preferably selected from the group consisting of boric acid and acetic acid/acetic salt. Most preferred is boric acid. Preferably the total concentration of the buffering agent (preferably of boric acid) in the composition is in the range from 5 g/L to 60 g/L, based on the total volume of the composition, preferably in the range from 10 g/L to 40 g/L, most preferably in the range from 20 g/L to 30 g/L.

As mentioned above, the present invention also refers to a method for electrolytically depositing a cobalt deposit onto a substrate, utilizing the composition of the present invention. If applicable, features of the aforementioned composition of the present invention apply likewise to the method of the present invention.

In step (A) of the method of the present invention, the substrate is provided. Preferred is a method of the present invention, wherein the substrate is a semiconductor base substrate. This means that the substrate preferably comprises at least one metalloid and/or gallium, more preferably selected from the group consisting of silicon, germanium and gallium, most preferably silicon.

Substrates for modern technical devices typically comprise features of small dimensions. Preferred is a method of the present invention, wherein the substrate comprises on at least one of its surfaces a plurality of vias and/or trenches.

More preferred is a method of the present invention, wherein the substrate is a wafer, more preferably a wafer with a plurality of vias and/or trenches on at least one of its surfaces.

A method of the present invention is preferred, wherein in step (C) the cobalt deposit fills said vias and/or trenches. More preferred is a method of the present invention, wherein in step (C) the cobalt deposit

- fills said vias and/or trenches and additionally

- at least covers the entire substrate on the side comprising said filled vias and/or trenches.

Subsequent to the deposition of the cobalt deposit, a method of the present invention is preferred, wherein in a further step the cobalt deposit is partly, horizontally removed, preferably by mechanical and/or chemical removal. As a result, a very smooth and uniform surface of the cobalt deposit is obtained.

Preferred is a method of the present invention, wherein the smallest opening dimension of said vias and trenches is 100 nm or less than 100 nm, preferably 50 nm or less than 50 nm, more preferably 30 nm or less than 30 nm, even more preferably 20 nm or less than 20 nm, most preferably 10 nm or less than 10 nm. Preferred is a method of the present invention, wherein said vias and trenches have an aspect ratio in the range from 2:1 to 50: 1 , preferably 3: 1 to 40: 1 , more preferably 3: 1 to 30: 1 , even more preferably 3: 1 to 20: 1 , most preferably 3: 1 to 15: 1.

Typically said substrate comprises a conductive seed layer for electrolytic deposition of cobalt. Preferred is a method of the present invention, wherein the substrate comprises a cobalt seed layer. Thus, in step (C) the cobalt deposit is preferably deposited onto a cobalt seed layer. This results in a very preferred homogeneous layer arrangement of seed layer and cobalt deposit. Usually such a homogeneous arrangement exhibits a very desirable resistivity. Furthermore, no additional barrier layer is preferably required.

In the method of the present invention, the substrate is operated as a cathode in order to obtain the cobalt deposit in step (C).

Preferred is a method of the present invention, wherein the electrical current is a direct current, preferably with a cathodic current density in the range from 0.01 A/dm² to 2 A/dm², more preferably with a cathodic current density in the range from 0.03 A/dm² to 1.5 A/dm², most preferably with a cathodic current density in the range from 0.05 A/dm² to 1.0 A/dm². In some cases it is preferred that the direct current in step (C) is not supplemented by current pulses. This means that preferably in step (C) the direct current is the only electrical current.

A method of the present invention is preferred, wherein in step (C) the contacting and supplying of electrical current is carried out for 3 seconds to 600 seconds, preferably for 5 seconds to 300 seconds, most preferably for 10 seconds to 150 seconds. If the contacting is carried out for significantly less than 3 seconds typically an incomplete cobalt deposit is obtained and often the plurality of vias and/or trenches is not or at least not sufficiently filled. If the contacting is carried out for significantly more than 600 seconds in most cases a too thick cobalt layer is deposited over the entire substrate, which is undesired for further processing steps, such as CMP, which are unnecessarily prolonged.

Preferably, a deposition sequence is carried out if direct current is applied. Therefore, a method of the present invention is preferred, wherein the contacting in step (C) comprises a first contacting time with a first cathodic current density, followed by a second contacting time with a second cathodic current density, wherein the first contacting time is shorter than the second contacting time and the first cathodic current density is lower than the second cathodic current density. Preferably, the first contacting time is in the range from 3 seconds to 90 seconds, more preferably from 5 seconds to 40 seconds. Independently from that, a preferred first ca- thodic current density is in the range from 0.01 A/dm² to 0.5 A/dm², more preferred in the range from 0.1 A/dm² to 0.4 A/dm².

Preferably, the second contacting time is in the range from 50 seconds to 510 seconds, more preferably from 70 seconds to 150 seconds. Independently from that, a preferred second cathodic current density is in the range from 0.1 A/dm² to 2.0 A/dm², more preferred in the range from 0.2 A/dm² to 1.0 A/dm².

By means of the electrical current the cobalt deposit is deposited on the substrate (preferably first into said vias and/or trenches and subsequently onto the entire substrate) in step (C) of the method of the present invention. Preferred is a method of the present invention, wherein the cobalt deposit comprises 60 wt.-% or more cobalt, based on the total weight of the cobalt deposit, preferably 75 wt.-% or more, more preferably 90 wt.-% or more, even more preferably 95 wt.-% or more, very much preferably 98 wt.-% or more, most preferably 99.9 wt.-% or more. A cobalt deposit comprising at least 99.9 wt.-% cobalt is usually considered as a pure cobalt deposit. Such a cobalt deposit is very much desired in modern electronic devices.

Very preferred is a method of the present invention, wherein the cobalt deposit is substantially free of (preferably does not contain) phosphorous and/or sulfur, preferably substantially free of (preferably does not contain) phosphorous and sulfur. Thus, the cobalt deposit is preferably a sulfur-free and/or phosphorous-free cobalt deposit. Furthermore, the cobalt deposit is preferably substantially free of (preferably does not contain) boron.

Preferred is a method of the present invention, wherein in step (C) the aqueous composition has a temperature in the range from 5°C to 90°C, preferably in the range from 15°C to 60°C, more preferably in the range from 20°C to 50°C, most preferably in the range from 22 °C to 30°C.

The present invention also refers to the use of at least one polymer comprising a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen in an aqueous composition of the present invention in order to deposit a uniform cobalt deposit onto a substrate. The meaning of "uniform" is explained in section 2.3 below in the text and in Fig. 9 to 11. Preferably, the use is additionally or alternatively to deposit a cobalt deposit all- over a substrate (i.e. without skip plating; the deposit is defect-free). More preferably, the use is to additionally fill (without voids) a plurality of vias and/or trenches in the substrate with the cobalt deposit from the bottom of said vias and trenches. If applicable, the aforementioned regarding the method of the present invention applies likewise to the use of said polymer.

The present invention is described in more detail by the following non limiting examples.

Examples

1. Aqueous compositions for depositing the cobalt deposit:

In a first step several aqueous compositions (No. C1 to C8 and E1 to E16) had been prepared. Each composition contained at least 90 wt.-% Dl water (based on the total weight of the composition), approximately 30 g/L boric acid and cobalt sulfate in such an amount that the total concentration of cobalt (II) ions in each composition was approximately 3 g/L (each based on the total volume of the composition). Table 1 summarizes the presence and concentration of further compounds. Each composition had a pH in the range between 4.0 and 4.5 and cobalt was the only transition metal in the compositions.

Table 1

First Compound Polymer

No. Label c [mg/L]* Label Mw [g/mol] c [mg/L]*

Comparative Examples

C1 (lb) 30

C2 (lb) 50

C3 (lb) 50 X1 5 000 000 200

C4 (lb) 50 X2 25 000 200

C5 (lb) 50 X3 1 000 000 100

C6 (lb) 50 X4 500 000 200

C7 (lb) 50 X5 354 5

C8 (lb) 50 X5 354 25

Examples according to the invention

E1 (lb) 30 (Ma) 8 000 200

E2 (lb) 50 (Ma) 8 000 200 E3 (lb) 30 (Ma) 8 000 400

E4 (lb) 50 (Ma) 8 000 400

E5 (lb) 30 (Ha) 1 300 000 200

E6 (lb) 50 (Ha) 1 300 000 200

E7 (lb) 30 (Ha) 1 300 000 400

E8 (lb) 50 (Ha) 1 300 000 400

E9 (lb) 30 (lid) 5 000 200

E10 (lb) 50 (lid) 5 000 200

E1 1 (lb) 30 (lid) 50 000 200

E12 (lb) 50 (lid) 50 000 200

E13 (lb) 30 (lid) 500 000 200

E14 (lb) 50 (lid) 500 000 200

E15 (lb) 30 (lie) 4 060 000 200

E16 (lb) 50 (lie) 4 060 000 200

* based on the total volume of the composition

In Table 1 , X1 , X2, X3, X4, and X5 have the following meaning:

X1 = Polyacrylamide (Sigma Aldrich)

X2 = Polyethylenimine (Sigma Aldrich)

X3 = Polyethylene glycol (Sigma Aldrich)

X4 = Poly(N-methylvinylamine) (Polyscience)

X5 = Bis-(sodium sulfopropyl)-disulfide, also abbreviated as SPS, which is not a polymer; these examples are for comparison reasons and are based on WO 2017/004424 A1. The specific molecular weight is 354 g/mol.

Compounds X1 , X2, X3, and X4 are polymers but not comprising a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen.

The utilized first compound (lb) with a molecular weight of 100.12 g/mol was purchased from BASF; polymer (I la) from Alfa Aesar; polymers (lie) and (lid), Mw 5 000 from Polyscience; (lid), Mw 50 000 and 500 000 from Sigma Aldrich. Similar results were also obtained with other first compounds comprising an acetylenic moiety (data not shown).

2. Method for electrolytically depositing a cobalt deposit onto a substrate (deposition, filling, and uniformity experiments):

In a second step the method of the present invention was carried out and sulfur-free cobalt was deposited and filled, respectively.

In each experiment, 500 ml of a respective aqueous composition prepared in the first step were used. For the sake of comparison, each number given in Table 1 (No.) refers to the corresponding experiment and example, respectively.

Prior to the experiments, each aqueous composition was purged for 15 minutes with inert nitrogen gas and was subsequently tested in view of (I) skip plating performance in respective deposition experiments, (II) filling performance in respective filling experiments, and (III) uniformity performance in respective uniformity experiments.

2.1 (I) skip plating performance (deposition experiments):

Skip plating refers to the completeness of the cobalt deposit and denotes undesired areas/spots within the cobalt deposit mostly free of deposited cobalt. In order to evaluate (I) skip plating performance, a blanket wafer (without features) was used as substrate and provided for each deposition experiment. On its active side, each substrate was equipped with a 3 nm TaN layer (deposited by PVD) and thereon a 10 nm cobalt layer (deposited by CVD) as conductive seed layer.

In each deposition experiment an electrical current with a current density of 0.6 A/dm² was supplied in order to electrolytically deposit the cobalt deposit, which was a layer having a thickness in the range from 220 nm to 280 nm. In each experiment, the substrate was contacted with the respective aqueous composition for approximately 100 seconds while said electrical current was supplied. During the experiments, the temperature of each composition was about 22°C.

Afterwards each substrate was rinsed with Dl water and dried by nitrogen gas flow such that a dried substrate was obtained.

Skip plating performance was evaluated for each dried substrate by visually inspecting images of each dried substrate. Imaging was carried out using a Laser Scanning Confocal Microscope (Olympus Lext OLS4100). The visual inspection was carried out by trained experts. Skip plating performance was categorized into the following four quality categories for skip plating (QCsp): A no or at least no significant skip plating is observable; the cobalt deposit is deposited all-over the substrate (very acceptable)

B few areas or a plurality of spots are not or insufficiently covered with cobalt (mostly acceptable)

C a significant area of the surface is not or insufficiently covered with cobalt (mostly not acceptable)

D the majority of the surface is not or insufficiently covered with cobalt (totally inacceptable)

The results are summarized in Table 2.

Comparative Examples C1 , C2, C7, and C8 are based on WO 2017/004424 A1 , wherein C1 and C2 do not contain any polymer or other additional organic additive. Example C1 shows an acceptable deposition quality (QCsp of B) wherein Example C2 is significantly worse (QCsp of C). According to Example C2, the skip plating performance decreases upon an increased concentration of the first compound comprising an acetylenic moiety.

Comparative Examples C7 and C8 additionally contain SPS, which is not a polymer. According to both examples, the presence of SPS does not positively affect the skip plating performance. In both examples an inacceptable QCsp of C was obtained.

Comparative Examples C3 to C6 additionally comprise a polymer, which does not comprise a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen. Polyacrylamide (Example C3) contains a plurality of carboxamide moieties each with a primary amine nitrogen. Polyethylenimine (Example C4) contains imine groups but no carboxamide moieties. Polyethylene glycol (Example C5) contains neither carboxamide moieties nor an amine nitrogen atom. Poly(N-methylvinylamine) (Example C6) contains a secondary amine nitrogen atom but no carboxamide moieties. Each of these examples shows an inacceptable skip plating performance (QCsp of C or D).

Table 2 shows that the skip plating performance can be significantly improved by utilizing at least one polymer comprising a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen in addition to the at least one first compound. Skip plaiting is in each experiment significantly reduced if such a polymer is utilized, irrespective of its molecular weight. This means that this positive result was observed in the presence of low molecular weight polymers (e.g. with a M_w of several thousand g/mol) as well as high molecular weight polymers (e.g. with a M_w of several million g/mol). As a result of the deposition experiments, in the presence of at least one polymer comprising a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen, the total concentration of the at least one first compound comprising an acetylenic moiety can be significantly increased without suffering undesired skip plating. This can, independent from these specific examples, very positively affect the final yield of deposited substrates, such as chips, because the rate of defective chips can be significantly reduced. It also means that an identical aqueous composition according to the present invention can be utilized in a very flexible way; it can be utilized for filling features with comparatively high aspect ratios but also for depositing cobalt layers without surface defects.

Since skip plating primarily occurs in the presence of comparatively high concentrations of said at least one first compound (e.g. 50 mg/L or more), deposition experiments are shown for aqueous solutions comprising a total concentration of 50 mg/L of first compound (lb). However, skip plating performance is reasonably expected to be identically good for all examples according to the present invention utilizing an aqueous composition comprising a total concentration of 30 mg/L of first compound (lb).

2.2 (II) bottom-up filling performance (filling experiments):

In order to evaluate said filling performance, wafers (Empire 1 M1 , SUNY Polytechnic Institute) were used as substrates and provided for each filling experiment. Each substrate has a 3 nm TaN layer (deposited by PVD), thereon a 4 nm cobalt seed layer (deposited by CVD) and was equipped with a plurality of trenches on its active surface with opening dimensions of approximately 100 nm and a depth of approximately 200 nm.

In each filling experiment an electrical current with a current density of 0.3 A/dm² was supplied in order to fill said trenches. Each substrate was contacted with the respective aqueous composition for approximately 30 seconds while said electrical current was supplied. During the experiments, the temperature of each composition was about 22°C.

Afterwards each substrate was also rinsed with Dl water and dried by nitrogen gas flow such that again dried substrates were obtained.

Bottom-up filling performance was evaluated for each dried substrate by visually inspecting vertical cross-sections of said filled trenches obtained after 5 second time intervals during said approximately 30 seconds (in particular after 10, 15, 20, 25, and 30 seconds). For this purpose FIB-SEM (FEI Helios Nanolab 450S) was carried out. Bottom-up filling performance was categorized into the following three quality categories for bottom-up filling (QCbuf): + no significant bottom-up filling; no formation of a horizontal deposition level is observed in the trenches (i.e. problematic for void-free filing of trenches with high aspect ratios)

++ significant bottom-up filling; a horizontal deposition level is observed in the fea- ture (mostly acceptable)

+++ explicit bottom-up filling; a clear horizontal deposition level is easily observed in the feature (very acceptable; i.e. provides high reliability in void-free filling of trenches with high aspect ratios)

The results are summarized in Table 2.

Table 2

No. QCsp QCbuf

Comparative Examples

C1 B +

C2 C +

C3 D

C4 D

C5 C

C6 D

C7 C

C8 C

Examples according to the invention

E1 +

E2 A +

E3 +

E4 A +

E5 +++

E6 A +++

E7 ++ E8 A ++

E9 +

E10 A +

E1 1 +

E12 A +

E13 +++

E14 A +++

E15 ++

E16 A +++ denotes not determined

Table 2 shows that in addition to an improved skip plating performance bottom-up filling performance can be significantly improved. However, this positive effect was primarily observed for comparatively high molecular weight polymers. Examples 13 and 14 indicate that on the one hand a molecular weight of 500 000 g/mol still facilitates this positive bot- ton-up filling performance, wherein on the other hand a molecular weight of more than 4 000 000 g/mol does in no way limit this positive effect (Examples 15 and 16).

An improved bottom-up filling performance is highly desired in order to fill features with high aspect ratios without voids.

According to Comparative Examples C1 and C2 no significant bottom-up filling performance is obtained in the trenches. Fig. 1 a shows the filling progress after 20 seconds for Example C1. Cobalt deposition primarily takes place on the side walls instead of at the bottom of the trenches; although finally a complete filling is obtained after 30 seconds (Fig. 1 b). However, this lack of bottom-up filling performance strongly indicates that filling of trenches with higher aspect ratios will result in undesired voids. Similar results were obtained in Comparative Example C2. After 10 seconds insufficient bottom-up filling is observed (Fig. 1 c) although again a complete filling is already obtained after 20 seconds (Fig. 1d).

In contrast to that, Example E5 (see Fig. 4a, after 15 seconds) and E6 (see Fig. 4b, after 15 seconds) show an explicit bottom-up filling by clearly showing a horizontal deposition level in the trenches. In these experiments filling of the trenches starts at the bottom and continues to the openings of the trenches, thereby avoiding the formation of undesired voids. This result was obtained with polymer (I la) having a Mw of 1 300 000 g/mol.

Similar results were obtained for Examples E13 and E14 (Fig. 6a and 6b, respectively, each after 15 seconds with polymer (lid), Mw of 500 000 g/mol), as well as Example E16 (Fig. 7, after 15 seconds with polymer (lie), Mw of 4 060 000 g/mol). In each case a clear horizontal deposition level was observed in the trenches.

Compositions according to the present invention but containing a polymer with a Mw significantly below 500 000 g/mol primarily showed an excellent skip plating performance without disturbing the filling of trenches. Example E2 (Fig. 2) and Example E4 (Fig. 3) show the filling progress after 30 seconds (polymer (I la), Mw of 8 000 g/mol). A similar result was obtained for Example E10 (Fig. 5, after 25 seconds; polymer (lid), M_w of 5 000 g/mol). In each case a filled trench was obtained.

2.3 (III) uniformity performance (uniformity experiments):

In order to evaluate said uniformity performance, substrates as used for above mentioned filling experiments (see 2.2 above) were used. Uniformity experiments were carried out for the following compositions as defined in Table 1 :

for comparison: C1 ,

according to the invention: E5, E9, and E15

In each uniformity experiment the substrate was contacted with the respective composition and an electrical current with a current density of 0.3 A/dm² for up to 30 seconds was supplied in order to fill the trenches. Subsequently, an electrical current with a current density of 0.6 A/dm² was supplied for approximately 100 seconds in order to electrolytically deposit a cobalt layer having a thickness in the range from 220 nm to 280 nm onto the substrate with filled trenches. During the experiments, the temperature of each composition was about 22°C.

Afterwards each substrate was rinsed with Dl water and dried by nitrogen gas flow such that again dried substrates were obtained.

Uniformity performance was evaluated for each dried substrate by visually inspecting vertical cross-sections of the dried substrates. For this purpose FIB-SEM (FEI Helios Nanolab 450S) was carried out.

Fig. 8 shows the result for the uniformity experiment with composition C1 (Comparative Example). As indicated by means of a black line (see "C" in Fig. 8) the cobalt deposit "B" is not uniformly distributed over the substrate "A" with its filled trenches. At the left end of the figure it is clearly seen that the layer thickness of the cobalt deposit is significantly lower compared to the layer thickness at the right end of the figure. In this case, uniformity performance is not satisfying. Generally, in subsequent processing steps such as CMP, such a cobalt deposit on a substrate typically causes problems and is therefore undesired.

On the contrary, Fig. 9 shows the result for the uniformity experiment with composition E5 (according to the invention). The cobalt deposit "B" is uniformly distributed over the substrate as indicated by black line "C". As seen over the entire figure, the layer thickness is approximately identical over areas with trenches and without trenches. In this case, uniformity performance is very satisfying and a subsequent CMP can be carried out without causing problems. A very similar uniformity performance was obtained with higher amounts of the first compound (lb) (data not shown) as well as with lower amounts of polymer (I la) (data also not shown).

Fig. 10 shows the result for the uniformity experiment with composition E9 (according to the invention). Again, the cobalt deposit "B" is uniformly distributed with approximately identical layer thickness over the substrate as indicated by black line "C". A very similar uniformity performance was obtained with higher amounts of the first compound (lb) (data not shown) as well as with lower amounts of polymer (lid) (data also not shown).

Fig. 1 1 shows the result for the uniformity experiment with composition E13 (according to the invention). Again, the cobalt deposit "B" is uniformly distributed with approximately identical layer thickness over the substrate as indicated by the black line "C". A very similar uniformity performance was obtained with higher amounts of the first compound (lb) (data not shown) as well as with lower amounts of polymer (lid) (data also not shown).

As shown in above uniformity experiments, the desired uniform layer thickness of the cobalt deposit is obtained with said at least one polymer with high molecular weight Mw (e.g. 1 300 000 g/mol) as well as with a low molecular weight Mw (e.g. 5 000 g/mol).

Claims

C L A I M S

An aqueous composition for depositing a cobalt deposit, the composition comprising

(a) cobalt (II) ions,

(b) at least one first compound comprising an acetylenic moiety,

(c) at least one polymer comprising a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen, and

(d) optionally a buffering agent.

The composition of claim 1 , wherein the total amount of said cobalt (II) ions in the composition represents 80 wt.-% to 100 wt.-% of all transition metal cations in the composition, based on the total weight of all transition metal cations in the composition, preferably at least 90 wt.-%, more preferably at least 95 wt.-%, even more preferably at least 98 wt.-%, most preferably at least 99 wt.-%.

The composition of claim 1 or 2, wherein the at least one first compound is represented by a compound of formula (I)

R¹ R²

(I),

wherein

R² denotes C1-C14 hydroxy alkyl or oxygen interrupted C2-C14 hydroxy alkyl. The composition of claim 3, wherein in Formula (I)

R denotes hydrogen, C1-C14 hydroxy alkyl, or -(CH₂)_n-0-R³,

R² denotes C1-C14 hydroxy alkyl or -(CH₂)_n-0-R³,

wherein independently in R and R²

n is in the range from 1 to 7 and R³ denotes C1-C7 hydroxy alkyl.

The composition of any of the preceding claims, wherein the acetylenic moiety in the at least one first compound is a terminal acetylenic moiety.

The composition of any of the preceding claims, wherein the at least one first compound is selected from the group consisting of (la) HC≡C-CH₂-OH,

(lb) HC≡C-CH₂-0-(CH₂)₂-OH, and

(lc) HC≡C-CH₂-0-(CH₂)₃-OH.

The composition of any of the preceding claims, wherein the total concentration of the at least one first compound in the composition is in the range from 5 mg/L to 250 mg/L, based on the total volume of the composition, preferably in the range from 7 mg/L to 150 mg/L, more preferably in the range from 9 mg/L to 90 mg/L, even more preferably in the range from 12 mg/L to 70 mg/L.

The composition of any of the preceding claims, wherein each of the at least one polymer has a weight average molecular weight Mw in the range from 1 200 g/mol to 50 000 000 g/mol, preferably in the range from 2 000 g/mol to 20 000 000 g/mol, more preferably in the range from 3 000 g/mol to 8 000 000 g/mol, even more preferably in the range from 4 000 g/mol to 6 000 000 g/mol, most preferably in the range from 5 000 g/mol to 5 000 000 g/mol.

The composition of any of the preceding claims, wherein the carboxamide moieties are independently represented by Formula (II)

(II),

wherein

R⁶ denotes a carbon atom of the backbone chain of the polymer.

The composition of any of the preceding claims, wherein the at least one polymer is selected from the group consisting of

(I la) Polyvinylpyrrolidone, (lib) Polyvinylcaprolactam,

(lie) Poly(2-methyl-2-oxazoline),

(lid) Poly(2-ethyl-2-oxazoline),

(lie) Poly(N-vinyl-acetamide), and

(I If) Poly(N-methyl-N-vinyl-acetamide).

11. The composition of any of the preceding claims, wherein the total concentration of the at least one polymer in the composition is in the range from 60 mg/L to 1000 mg/L, based on the total volume of the composition, preferably in the range from 80 mg/L to 800 mg/L, more preferably in the range from 100 mg/L to 600 mg/L, even more preferably in the range from 150 mg/L to 500 mg/L, most preferably in the range from 200 mg/L to 400 mg/L.

12. A method for electrolytically depositing a cobalt deposit onto a substrate, the method comprising the steps

(A) providing the substrate,

(B) providing an aqueous composition according to any of claims 1 to 11 ,

13. The method of claim 12, wherein the substrate comprises on at least one of its surfaces a plurality of vias and/or trenches.

14. The method of claim 13, wherein in step (C) the cobalt deposit fills said vias and/or trenches.

15. Use of at least one polymer comprising a plurality of carboxamide moieties each with a secondary or tertiary amine nitrogen in an aqueous composition as defined in one of claims 1 to 1 1 in order to deposit a uniform cobalt deposit onto a substrate.