EP2959524A1

EP2959524A1 - Carbon nanotube-containing dispersion and the use thereof in the production of electrodes

Info

Publication number: EP2959524A1
Application number: EP14705357.3A
Authority: EP
Inventors: Dagmar Ulbrich; Werner Hoheisel; Joachim Ritter; Lars Krueger
Original assignee: Covestro Deutschland AG
Current assignee: Covestro Deutschland AG
Priority date: 2013-02-22
Filing date: 2014-02-20
Publication date: 2015-12-30
Also published as: WO2014128190A1; DE102013213273A1; CN105074966A; US20160020466A1; JP2016514080A; KR20150122653A

Abstract

The invention relates to a dispersion comprising a dispersion medium, a polymeric dispersing agent, and carbon nanotubes dispersed in the dispersion medium. The proportion of carbon nanotubes present in the form of agglomerates with an average agglomerate size of ≥ 1 μm to the total quantity of carbon nanotubes ≤ 10 vol%, and ≥ 70% of the carbon nanotubes which are not present in the agglomerated form have a length of ≥ 200 nm. the invention further relates to a method for producing such a dispersion, to a method for producing an electrode with the dispersion, to an electrode obtained in this manner, and to an electrochemical element containing the electrode.

Description

- i -

Carbon nanotube-containing dispersion and its use in the manufacture of electrodes

The present invention relates to a dispersion comprising a dispersion medium, a dispersing aid and carbon nanotubes dispersed in the dispersion medium. It also relates to a slurry comprising the dispersion and an active material which is common for secondary batteries and which is capable of either removing or storing lithium ions during the discharging and charging process, depending on the use in the positive or negative electrode, and optionally further additives. In addition, it relates to a positive or negative electrode containing the dispersion or slurry according to the invention. It further relates to a process for the preparation of such a dispersion, a process for the preparation of such a slurry, a process for the preparation of an electrode with the dispersion or slurry, an electrode obtained therefrom and an electrochemical element containing the electrode.

Carbon nanotubes (CNTs) are known for their exceptional properties. For example, their tensile strength is about 100 times that of steel (eg ST52), whose thermal conductivity is about as high as that of diamond, their thermal stability reaches up to 2800 ° C in vacuum and their electrical conductivity can be many times the conductivity amount of copper. However, these structure-related characteristics are often only accessible at the molecular level if it is possible to homogeneously distribute carbon nanotubes and to produce the largest possible contact between the tubes and the medium, ie to make them compatible with the medium and thus stable dispersible. With regard to electrical conductivity, it is furthermore necessary to form a network of tubes in which, ideally, they only touch one another at the ends or approach sufficiently closely. Here, the carbon nanotubes should be as isolated as possible, that is agglomerate-free, and not aligned. In this case, the carbon nanotubes can be present in a concentration at which such a network can just form, which is reflected by the sudden increase in electrical conductivity as a function of the concentration of carbon nanotubes (percolation limit).

The addition of CNTs as a conductive additive to the electrode material for lithium ion batteries and accumulators is of particular interest here. For this purpose, the CNTs should be dispersed in a preferred dispersion medium using as little dispersant as possible. In addition, for industrially relevant applications, the highest possible concentration of CNTs in the dispersion should be ensured, which is in the range of well over 1% by weight and preferably above 3% by weight, more preferably above 4% by weight. It has been shown that for the performance of such an electrode not only the quality of the dispersion in terms of intrinsic morphological Characteristics of CNTs such as length, aspect ratio, surface or defect density but also the degree of dispersion or agglomerate content is important. In the latter case, the CNTs have the ability to cluster homogeneously around the individual particles of the active material. Furthermore, the type and amount of the dispersing assistant have an influence on the electrical properties, in particular on the impedance, of the CNT network and thus also of the electrode as a whole.

A high-performance electrode is characterized by high power and energy density as well as by a long service life or Zyklisierbarkeit. High power densities in a battery are achieved in particular by a high conductivity of the electrode, for which a good wetting of the active material is essential. Overall, the individual components of the electrochemical impedance should be as low as possible, including minimum contact resistance between well-dispersed CNTs with each other and the Abieiter the electrode. For this reason, sufficient stabilization should be achieved even with the smallest possible amount of electrically insulating dispersing agent in the dispersion, but should also have a sufficiently long shelf life of at least several months.

In the prior art, for example, US 2007/224106 Al deals with CNT dispersions. Here, the function of a nonionic surfactant for the dispersion of CNTs will be described. According to this patent application, it has been found that a mixture of an amide-based organic solvent and a polyvinylpyrrolidone or an amide-based organic solvent, a nonionic surfactant and a polyvinylpyrrolidone can disperse CNTs well. Ultrasonic treatment is described as necessary to disperse the CNTs. The ultrasonic treatment may be performed during the dispersion step of the CNTs in the nonionic surfactant and / or the amide-based polar organic solvent. Alternatively, a mixture of the nonionic surfactant and / or the amide-based polar organic solvent and polyvinylpyrrolidone can be prepared, and the sonication is performed during the dispersion of the CNTs herein.

A disadvantage of the process described in US 2007/224106 Al is that an ultrasound-based dispersion that lasts for many minutes or hours can not be applied or only with great effort as an industrial process and that this method, as known in the art , is only applicable to dispersions with low concentrations and low viscosities. Moreover, ultrasonic treatments of CNTs of nanotubes breaks, leaving less of the desired CNTs with a high length to diameter ratio.

DE 10 2005 043 054 A1 (WO 2007/028369 A1) relates to a dispersion consisting of a dispersing liquid and at least one solid which is distributed in the dispersing liquid, wherein the dispersing liquid has an aqueous and / or non-aqueous base, which is at least one solid made of graphite and / or porous carbon and / or carbon nanomaterial and / or coke and which distributes at least one solid homogeneously and stably in the dispersing liquid is. In example 3 of this patent application, 10 g of carbon nanotubes (CNTMW) without additive additive are dispersed in 500 ml of 2-propanol. The carbon nanotubes had a diameter of 10-20 nm and lengths of 1-10 μιη and their BET specific surface area was 200 m ² / g. The predispersion with a viscosity of 600 mPa s was exposed to a shear rate of 2 500 000 / sec at a pressure of 1000 bar. As a separate repetition of this experiment has shown, however, are still getting too large particles.

An example of CNT-containing electrodes is the patent application WO 2012/1 14590. It deals with an electrode for a non-aqueous electrolyte based secondary battery. The electrode comprises an active material, a binder, CNTs and a non-fibrous conductive carbon material, wherein a PVP-based polymer is contained in a proportion of 5 to 25 parts by weight of 100 parts by weight of CNTs.

The object of the present invention has been to at least partially overcome the disadvantages of the prior art. In particular, it has set itself the task of providing CNT dispersions which can be used in the manufacture of improved electrodes for batteries and accumulators or supercapacitors. A process for preparing such dispersions is also an object of the invention.

According to the invention, this object is achieved by a dispersion comprising a dispersion medium, a preferably polymeric dispersant and carbon nanotubes dispersed in the dispersion medium, the proportion of carbon nanotubes present in agglomerates with an average agglomerate size of> 1 μιη in the total amount of carbon nanotubes 40% by volume and> 70% by weight of the non-agglomerated carbon nanotubes have a length of> 200 nm.

The dispersions of the invention can be used to prepare electrodes in a lithium ion battery with increased power density and extended life. Secondary products are a slurry for application to an electrode conductor and the provision of the electrode for a lithium-ion battery. In other words, the dispersion may serve as a base for preparing a slurry applied to a suitable current collector (preferably aluminum for the positive electrode and copper for the negative electrode), drying and calendering for producing a battery or accumulator electrode. Preferably, the proportion of carbon nanotubes present in agglomerates with an average agglomerate size of> 1 μm in the total amount of carbon nanotubes is <20% by volume, more preferably <10% by volume. The term "volume%" in the following refers to the cumulative volume-related cumulative distribution Q3 known to the person skilled in the art, which describes an upper or lower range or an interval of the corresponding distribution. The volume percentages described here refer to values which are determined with a laser diffraction device for measuring a particle size distribution.

It is furthermore preferred that> 80% by weight, more preferably> 90% by weight, of the carbon nanotubes present in unagglomerated form have a length of> 200 nm. This can be determined by means of transmission electron microsopy of a corresponding dispersion sample. Of course, a few contacts between individual CNTs do not mean that the CNTs have to be labeled as agglomerated.

Carbon nanotubes (CNTs) within the meaning of the invention are all single-walled single-walled or multi-walled carbon nanotubes of the cylinder type (for example in patent Iijima US 5,747,161; Tennant WO 86/03455), scroll type, multiscroll type, cup-stacked type closed or open on both sides, conical cups consisting (for example in the patent Geus EP198,558 and Endo US 7018601B2), or onion-like structure. Preference is given to using multi-walled carbon nanotubes of the cylinder type, scroll type, multiscroll type and cup-stacked type or mixtures thereof. It is favorable if the carbon nanotubes have a ratio of length to outer diameter of> 5, preferably> 100.

In contrast to the known scroll-type carbon nanotubes with only one of the above-mentioned graphite molecules, carbon nanotube structures which consist of several graphene layers that form a single layer of carbon nanotubes are also known Stack summarized and rolled up. This is called the multiscroll type. These carbon nanotubes are described in DE 10 2007 044031 AI, to which reference is made in its entirety. This structure is similar to the simple scroll-type carbon nanotubes as compared to the structure of multi-walled cylindrical carbon nanotubes (cylindrical MWNT) to the structure of single-walled cylindrical carbon nanotubes (cylindrical SWNT).

Unlike the onion type structures, the individual graphene or graphite layers in these carbon nanotubes, seen in cross-section, evidently run continuously from the center of the carbon nanotubes to the outer edge without interruption. Preferably, but not exclusively, the viscosity of the dispersions according to the invention is adjusted by varying the CNT concentration and not via the dispersing aid. Also, preferably, but not exclusively, the viscosity in a window at a shear rate of 1 / s should be between about 0.01 Pa-s and about 1000 Pa-s, preferably between 0.1 Pa-s and about 500 Pa -s and more preferably between 1 Pa-s and about 200 Pa-s to ensure good processability of the dispersion and the slurries derived therefrom to electrode layers with suitable layer thicknesses. The viscosity can be measured with a suitable rotary viscometer (eg Fa. Anton Paar, MCR series).

Embodiments and other aspects of the invention are described below. They can be combined with each other as long as the context does not clearly indicate the opposite.

In one embodiment of the dispersion according to the invention, the dispersion medium is selected from the group consisting of water, acetone, nitriles, alcohols, dimethylformamide (DMF), N-methyl-2-pyrrolidone (NMP), pyrrolidone derivatives, butyl acetate, methoxypropyl acetate, alkylbenzenes, cyclohexane derivatives and mixtures hereof. Preference is given to using water, dimethylformamide (DMF), N-methyl-2-pyrrolidone (NMP) and / or pyrrolidone derivatives.

In a further embodiment of the dispersion according to the invention, the dispersing assistant is selected from the group consisting of poly (vinylpyrrolidone) (PVP), polyvinylpyridines (eg poly (4-vinylpyridine) or poly (2-vinylpyridine)), polystyrene (PS), poly (4-vinylpyridine -co-styrene), poly (styrenesulfonate) (PSS), lignin sulfonic acid, lignosulfonate, poly (phenylacetylene) (PPA), poly (meta-phenylenevinylene) (PmPV), polypyrrole (PPy), poly (p-phenylenebenzobisoxazole) (PBO) , naturally occurring polymers, anionic aliphatic surfactants, poly (vinyl alcohol) (PVA), polyoxyethylene surfactants, poly (vinylidene fluoride) (PVdF), cellulose derivatives (in general, and in particular those in which the hydrogen atom of some hydroxyl groups on the glucose units is methyl or ethyl). or higher groups have been replaced, such as methyl cellulose (MC) or ethyl cellulose (EC), cellulose derivatives in which the hydrogen atom of some hydroxyl groups on the glucose units by hydroxymethyl, hydroxyethyl, hydroxypropy L- or higher groups have been replaced, such as hydroxymethylcellulose (HMC) hydroxyethylcellulose (HEC) or hydroxypropylcellulose (HPC), cellulose derivatives in which the hydrogen atom of some hydroxyl groups on the glucose units have been replaced by carboxymethyl, carboxyethyl or higher groups, such as carboxymethylcellulose ( CMC) or carboxyethylcellulose (CEC), cellulose derivatives in which the hydrogen atom of some hydroxyl groups on the glucose units have been partially replaced by alkyl groups and partly by hydroxyalkyl groups, such as hydroxyethyl methyl cellulose (HEMC) or hydroxypropyl methyl cellulose (HPMC)), mixtures of different cellulose derivatives, Polyacrylic acid (PAA), polyvinyl chloride (PVC), polysaccharides, styrene-butadiene rubber (SBR), polyamides, polyimides, block copolymers (for example, acrylic block copolymers, ethylene oxide-propylene oxide copolymers), and mixtures thereof.

Other biocidal additives may be added as needed. These then do not act as dispersants themselves, but contribute to the durability of the dispersion when these bacteria, fungi, yeast or algae colonizing natural substances such. Contain celluloses and their derivatives or lignin sulfonic acid as a dispersing aid.

Many dispersants are ionic in nature and contain sodium as counter-ion. In a further embodiment, the dispersing aid comprises lithium ions. These lithium ions as counter-ions can be introduced, for example, either directly during production or later exchanged with the aid of ion exchangers. By way of example, mention may be made here of carboxymethylcellulose (CMC) or polyacrylic acid (PAA). The advantage of this approach is in a presaturation of the material, so that in subsequent use in electrodes often occurring during the first charge and discharge cycle capacity loss can be excluded or at least decisively reduced.

Incidentally, the combination of N-methyl-2-pyrrolidone (NMP) as dispersion medium together with PVP, EC, MC, polyvinylpyridine, poly (4-vinylpyridine-co-styrene), polystyrene (PS) or mixtures thereof as dispersion auxiliaries is expressly preferred , In addition, it is expressly preferred to use the combination of water as the dispersing medium and PVP or cellulose derivatives such as e.g. with CMC (or SBR instead of PVP) or mixtures of the two as a dispersing aid.

Also preferred are low molecular weight types of these dispersants, wherein PVP has a number average molecular weight less than 200,000 g / mol, more preferably between 10,000 g / mol and 100,000 g / mol, most preferably between 25,000 g / mol and 75,000 g / mol. Likewise preferred are those which have a low viscosity in order to be able to produce a higher CNT concentration. When CMC is used as dispersing assistant, not too high degrees of substitution are preferred, which should be between 0.5 and 1.5, preferably between 0.6 and 1.1, in order to stabilize the dispersion well by, on the one hand, good affinity for polar media how to obtain water and on the other hand to ensure a stable non-covalent attachment to the CNT by sufficiently hydrophobic moieties in the CMC molecule.

In a further embodiment of the dispersion according to the invention, the carbon nanotubes present in non-agglomerated form are multi-walled carbon nanotubes with an average outer diameter of> 3 nm to <100 nm, preferably> 5 nm to <50 nm and a length to diameter ratio of> 5, preferably> 100. In a further embodiment of the dispersion according to the invention, the carbon nanotubes are present in a proportion of> 1% by weight and <25% by weight, preferably> 3% by weight and <15% by weight, based on the total weight of the dispersion.

In a further embodiment of the dispersion according to the invention, the ratio of the concentration of the dispersing aid in the dispersion medium and the concentration of the carbon nanotubes in the dispersion medium is in a range of> 0.01: 1 to <10: 1, preferably> 0.01: 1 to <0, 9: 1, more preferably> 0.01: 1 to <0.6: 1, very particularly preferably> 0.02: 1 to <0.3: 1. The smallest possible proportion of dispersing aids is preferred here in order to minimize any interfering influence of these aids in the later application.

In a further embodiment of the dispersion according to the invention, this further comprises conductive carbon black, graphite and / or graphene. Preferably, the mass ratio of carbon nanotubes and at least one element of these classes of materials is between 1:10 and 10: 1, and more preferably between 1: 3 and 3: 1. Furthermore, at least one element of this material class can also be added during the preparation of the electrode slurry (see below) as a separate dispersion or as a powder. The benefit of adding such carbonaceous conductive materials has been found empirically and is believed to be due to a better pore structure of the electrode. Incidentally, this can achieve a cost savings.

In a more differentiated view, the specific surface area of the CNTs can be related to the relative proportion of dispersing agent. With increasing specific surface area (according to Brunauer, Emmett, Teller: BET) of the CNTs, the relative proportion of dispersing agent to CNTs should increase steadily. Thus, for CNTs with a specific surface area (according to BET) of about 130 m ² / g (for example Baytubes C70P, Bayer AG) in the dispersion medium NMP and PVP as dispersing aids (for example PVP K30, Luvitec, BASF AG) for the concentration ratios of PVP and CNT: 0.01: 1 <(C _PV P: CCNT) <0.5: 1, preferably 0.02: 1 <(CPVP: CCNT) <0.25: 1, more preferably 0.04: 1 <(CPVP: CCNT) <0.2: 1, most preferably 0.06: 1 <(CPVP: CCNT) <0.18: 1. where C _PV p or CCNT is the concentration (% by weight) of PVP or CNT in the dispersing medium. For CNT with a specific surface area (according to BET) of about 210 m ² / g (for example Baytubes C150P, Bayer AG), NMP for the concentration ratios of PVP (for example PVP K30, Luvitec, BASF AG) and CNT applies in the dispersion medium: 0.02: 1 <(CPVP: CCNT) <0.6: 1, preferably 0.06: 1 <(C _PV P: CCNT) <0.4: 1, more preferably 0.1: 1 <(CPVP: CCNT) <0.3: 1, most preferably 0.15: 1 <(CPVP: CCNT) <0.25: 1.

When using CNTs with a specific surface area (according to BET) of about 130 m ² / g (for example Baytubes C70P, Bayer AG) in the dispersion medium NMP and ethylcellulose as dispersing agent (for example ETHOCELL 100, Dow Wolff Cellulosics) similar concentration ratios of Ethylcellulose (EC) and CNT: 0.01: 1 <(CEC: CCNT) <0.5: 1, preferably 0.02: 1 <(CEC: CCNT) <0.25: 1, more preferably 0.04: 1 <(CEC: CCNT) <0.2: 1. Where Cpw or CEC is the concentration (by weight) of %) of EC or CNT in the dispersing medium. When methylcellulose, other cellulose derivatives, polystyrene, poly (4-vinylpyridine) or poly (4-vinylpyridine-co-styrene) are used, similar concentration ratios apply to those CNTs having a similar BET specific surface area as Baytubes C70P (Bayer AG ). When using CNT with a higher specific surface area (according to BET), such as Baytubes C150P (Bayer AG) with a specific surface area (according to BET) of approx. 210 m ² / g or Nanocyl 7000 with a specific surface area (according to BET) of 250 to 300 m ² / g, correspondingly higher concentrations of dispersing agent must be used, as was already carried out using the example of PVP.

The same concentration ratios also apply when water is used as the dispersing medium and CMC (for example Walocel CRT 30G, Dow Chemicals) is used as the dispersing aid.

Another aspect of the present invention is a process for producing a dispersion according to the invention wherein a precursor dispersion comprising a dispersion medium, a polymeric dispersing aid and carbon nanotubes is dispersed by means of a high pressure homogenizer.

In the preparation of the dispersions of the invention, pretreatments of CNTs, which may also consist of commercially available materials (for example Baytubes C70P or C150P, Nanocyl NC7000 from Nanocyl S.A. or AMC from ÜBE Industries), are optional. Depending on the moisture content of the CNTs may optionally be followed by drying (preferably 60-150 ° C for 30 - 150 min) carried out in air.

Subsequently, also optionally, a pre-crushing of large CNT agglomerates in a manner that does not change the morphological structure of the CNT (tube structure is retained) except for a certain shortening. The d50 value (laser diffraction) of the agglomerate size after pre-shredding is, for example, <100 μm, preferably <30 μm, particularly preferably <10 μm.

As a method of the pretreatment, preferred is dry milling, which may be carried out by means of a knife mill, mortar mill, planetary ball mill or other suitable mill known to the expert. The purpose of this process step is to provide smaller, more compact CNT agglomerates, and is optionally used only to optionally prevent blockages of nozzles, lines or valves as used in one or more of the following steps. Whether a pre-treatment with a mill is necessary depends on the morphology of the used CNT agglomerates and on the effectiveness of the subsequent predispersion process. For predispersion, a mixture of the CNT powder with dispersing aid and dispersion medium having the desired concentration and viscosity is prepared. The mixing process takes place, for example, with a disperser with high shear forces, such as a rotor-stator system, until a homogeneous dispersion with dispersion medium, dispersing agents and CNT agglomerates (one size (d50, laser diffraction) of <500 μm, preferably <100 μm, is particularly preferably have <50 μιη) is present. Corresponding apparatus with a rotor-stator system are offered, for example, by Fluid Kotthoff GmbH, Germany, or Cavitron GmbH, Germany.

Thereafter, the dispersion is carried out by means of a high-pressure homogenizer. These consist essentially of a high-pressure pump and at least one nozzle for the homogenization. In this case, the pressure built up by the high-pressure pump is released in the homogenizing valve, which causes the dispersion of the CNT agglomerates. High-pressure systems with which dispersions of CNTs can be prepared are described with regard to the principle of e.g. in S. Schultz et al. (High-Pressure Homogenization as a Process for Emulsion Formation, Chem. Eng. Technol. 27, 2004, pp. 361-368). A special design of the high-pressure homogenizer is that of the jet disperser. In this case, a high pressure is built up with a pump, which is expanded by a circular, gap or otherwise shaped nozzle. Without limiting the generality, the pump can continuously, but also discontinuously z. B. pneumatically or hydraulically via piston operation, operated. The nozzle can be equipped with a single opening or bore. However, it is also possible to use nozzles with two or more openings, which are respectively arranged opposite one another or in the shape of a ring (see, for example, DE 19536845). In a preferred embodiment, the nozzles are made of non-ferrous ceramic materials such as aluminum oxides, which optionally also zirconium, yttrium or other common for ceramics oxides are added, or made of other metal carbides or nitrides. The advantage of using these materials is that contamination of the dispersion with iron is avoided. Jet dispersants in general are described, for example, in Chemie Ingenieur Technik, Volume 77, Issue 3 (pp. 258-262).

In principle, the fineness of the dispersion produced depends on the pressure and the nozzle used. The smaller the nozzle bore or gap width and the higher the pressure, the finer the dispersion obtained. Smaller orifices generally require and allow higher working pressures. However, too small nozzle bores can lead to blockages or impose excessive limitations on the applicable viscosities, which in turn limits the usable CNT concentrations. Too high a pressure can also permanently damage the morphology or structure of the CNTs, so that an optimum in the apparatus parameters must be found for each system comprising dispersion medium CNT and dispersing assistant. The pressure difference Δρ is, for example, Δρ> 50 bar, preferably Δρ> 150 bar, particularly preferably 1500 bar>Δρ> 250 bar, and very particularly preferably 1200 bar>Δρ> 500 bar. In principle, smaller bore diameters or gap widths lead to better dispersing results (ie higher fraction of isolated CNT in the dispersion), although the risk increases that larger agglomerates block or clog the nozzle (s). Two or more opposing nozzles have the advantage that the abrasion in the nozzle is minimized, since the dispersing jet is not directed against a solid baffle plate and thus entrained impurities in the dispersion are minimized.

The viscosity of the dispersion also limits the choice of the bore diameter or gap width downwards. Depending on the agglomerate size and viscosity of the dispersion, suitable nozzle diameters therefore have to be adapted in a manner known to the person skilled in the art.

Another preferred high-pressure homogenizer for producing the dispersion according to the invention is characterized by a valve with a variable width of a gap. In this case, a pressure is built up by a pump in a volume which releases a gap via a movable plunger, via which the dispersion is expanded by the pressure gradient. In this case, the gap width and thus the pressure built up manually, can be adjusted via a mechanical or electrical control circuit. However, the gap width or the built-up pressure can also be automatically regulated via the counterforce of the punch, which can be adjusted, for example, via a spring. The gap is often an annular gap. The process is also tolerant to agglomerates in the range of 100 μιη. The corresponding process is known, described in EP0810025 and corresponding devices are marketed, for example, by GEA Niro Soavi (Parma, Italy)

Another preferred high-pressure homogenizer for preparing the dispersion of the invention operates discontinuously and compresses the dispersion via a die in a piston cylinder with> 500 bar, preferably> 1000 bar. The relaxation of the dispersion takes place via a gap, preferably via an annular gap. The method is known and corresponding devices are marketed, for example, by the company APV Gaulin Deutschland GmbH, Lübeck, Germany (for example Micron LAB 40).

In one embodiment of the process according to the invention, the dispersion is carried out several times by means of the high-pressure homogenizer. The dispersing process can thus be repeated until a satisfactory separation of the CNTs has taken place. The number of repetitions depends on the material used, CNT concentration, viscosity and applied pressure and can be 30, 60 or even more than 100 times. In general, the number of passes required increases with the viscosity of the dispersion. From an economic point of view, an upper limit from the point of view of dispersion quality is sensible, but not technically required because the quality of the dispersion decreases only too slowly due to too frequent repetition

Another aspect of the dispersion result is not only the total energy that is introduced into the dispersion but also the power density or stress intensity (energy input per time and volume of the dispersion) of the CNT agglomerates in the dispersion. This means that when falling below a certain pressure difference, a fine distribution is difficult, regardless of the total energy input. The lower limit of the power density required to obtain a good dispersion result is product-specific and depends on the type of CNT, its pretreatment, the solvent and the dispersing aids. The necessary total energy based on the amount of CNT used to achieve a good dispersion results when using small pressure differences (about 200 bar) about 40,000 kJ / kg, at high pressure differences of over 800 bar and less than 15,000 kJ / kg.

From an economic point of view, it usually makes sense to limit this to around 150 repetitions. However, a higher number is not exclusive for the process according to the invention. The repetitions can be staggered at the same nozzle according to a driving style in a circle. But they can also be spatially offset on nozzles that are connected in series or in a combination of limited in the number, spatially offset nozzles and a driving in a circle. A method according to the invention is also present if the repetitions are divided into blocks in such a way that the dispersing operations in the individual blocks take place with different nozzle sizes, nozzle shapes and working pressures. This may be particularly recommendable if the viscosity of the dispersion changes during the dispersion. Advantageously, the use of initially larger and later, when the viscosity decreases, to smaller bore diameters or gap widths. The use of continuously adjustable gap widths, for example as described in DE 10 2007 014487 A1, offers advantages. During the dispersing process of CNTs, the transition from larger agglomerates to individual CNT fibers is often accompanied by an increased viscosity, which obstructs or even makes passage through the nozzles impossible, ie leads to blockages. For this reason, in a further embodiment of the process according to the invention, first of all a series of predispensed mixtures with an increasing concentration of CNTs in the dispersion medium and the appropriate amount of dispersing assistant are prepared. Thereafter, the pre-spray mixtures are fed consecutively, starting from the lowest concentration, to the high-pressure homogenizer. After the addition of the last, most highly concentrated mixture, a total dispersion is then obtained with an average concentration based on the predispersed mixtures. In a further embodiment, a lower-concentration partial dispersion, which has already been subjected to a treatment with the high-pressure homogenizer, is added by addition of CNT Powder, which was optionally already supplied to a pre-crushing, concentrated and fed to a new predispersion. This mixture is then treated again with the high-pressure homogenizer and gives a dispersion having a relation to the first dispersion increased concentration. If necessary, this process can be repeated until the desired final concentration of the dispersion and total amount has been reached.

In a further embodiment of the method according to the invention, the high-pressure homogenizer is a jet disperser and has at least one nozzle with a bore diameter of> 0.05 to <1 mm and a length to diameter ratio of the bore of> 1 to <10, wherein between the nozzle inlet and Nozzle outlet is a pressure difference of> 5 bar.

In a further embodiment of the method according to the invention, the jet disperser has at least one slot gap with a gap width of> 0.05 to <1 mm and a depth to gap width ratio of the slot gap of> 1 to <10, wherein between the nozzle inlet and nozzle outlet a pressure difference of > 5 bar. A preferred jet disperser for the preparation of the dispersion according to the invention is described in DE 19536845 Al. The bore diameter or gap width in the nozzle for the present invention is, for example, 0.1 mm to 1 mm, preferably from 0.2 mm to 0.6 mm. Further refinements of the jet dispersant are described, for example, in DE 10 2007 014487 A1 and WO 2006/136292 A1. The present invention further relates to a composition for producing an electrode comprising a dispersion according to the invention, an electrode material and a polymeric binder, wherein the binder is present at least partially in dissolved form in the composition. Optionally, particulate graphite or conductive carbon black may be added as a conductive material to the composition. The mentioned "composition" is also called slurry. The preparation of the slurry is carried out by mixing the dispersion of the invention, a suitable binder which dissolves or disperses in the dispersion medium and an active material for intercalation and storage of lithium ions. It is preferred to pay attention to the setting of a suitable viscosity with the highest possible solids content. For the electrode material, the known classes of materials can be used. For positive electrodes, lithium intercalating compounds such as layered compounds, spinels or olivines can be used. a preferred embodiment, the composition is the electrode material selected from the group LiNi _x Mn _y Al _z COI _x - _y - _z 02 (0 <x, y, z <l and x + y + z <l) LiNi ₀ , 33Mno, 33Coo, 330 ₂ , LiCoO ₂ , LiNi ₀ , 7Coo, 30 ₂ , LiNi ₀ , 8Coo, ₂ 0 ₂ , LiNi ₀ , 9Coo, i0 ₂ , LiNiO ₂ , LiMn ₂ 0 ₄ , LiMni. ₅ (Co, Fe, Cr) o, ₅ 04, L i N i _x Al _y Co _x - _y 0 ₂ (0 <x, y <l and x + y <l), LiNio, 8Coo, i ₅ Alo , o50 ₂ , LiNio, 78Coo, i9Al ₀ , o30 ₂ , LiNi ₀ , 78Coo, i9Al ₀ , o3M _x 0 ₂ (x = 0.0001-0.05, M = alkali or alkaline earth metals), LiFeP0 ₄ , Li ₂ FeP ₂ 0 ₇ , L1C0PO4, Lii _{+ x} M _y Mn ₂ . _x . _y 0 ₄ (M = Al, Cr, Ga), LiTiS ₂ , Li ₂ V ₂ 0 ₅ , LiV ₃ O ₈ , L T1S3, Li3NbSe3, L T1O3, sulfur, polysulfides and / or sulfur-containing materials. The materials may be present as micro- or nanoparticles. With such materials, positive electrodes can be realized.

In a further preferred embodiment of the composition, the electrode material is selected from the group of natural or synthetic graphite, hard carbon, which has a stable random structure of interconnected very small and thin carbon platelets, softer, (substantially) graphitic carbon, silicon, silicon alloys , Silicon-containing mixtures, lithium titanate (Li ₂ TiÜ 3 or Li 4 TisOi ₂ ), tin alloys, C03O4, Li ₂ , 6Coo, 4N and / or tin oxide (SnO ₂ ). The materials may be present as micro- or nanoparticles. With such materials, negative electrodes can be realized. It is also preferred that the binder is selected from the group poly (vinylidene fluoride) (PVdF), carboxymethylcelluloses (CMC), types of butadiene rubber such as styrene-butadiene rubber (SBR), acrylonitrile-butadiene rubber (acrylonitrile-butadiene rubber), polyacrylic acid or combinations thereof.

The combination of NMP as dispersion medium with PVP, ethylcellulose, methylcellulose, polyvinylpyridine, polystyrene or polyvinylpyridine-polystyrene block copolymers as dispersing assistant and PVdF as binder is expressly preferred.

Even more preferred is the combination of NMP as the dispersion medium with PVP, ethylcellulose, methylcellulose, polyvinylpyridine, polystyrene or polyvinylpyridine-polystyrene block copolymers as a dispersing aid, PVdF as a binder and NMC as an electrode material.

Also preferred is the combination of water as the dispersion medium with CMC or PVP as a dispersing aid and SBR or polyacrylic acid as a binder. Some of the materials mentioned have an effect both as dispersing aids and as binders. Thus, in this invention, a material should not be limited to only one of these functions. For example, polyacrylic acid has an effect as a dispersing aid and is at the same time used as a binder for anodes in Li-ion batteries for example (see eg A. Magasinski et al., ACS Appl Mater Interfaces, 2010 Nov; 2 (l 1): 3004 -10th doi: 10.1021 / aml00871y). According to the invention, a combination of polyacrylic acid and CMC, in which small amounts of CMC are used, greatly improve the stabilization over the use of pure polyacrylic acid. The present invention furthermore relates to a method for producing an electrode, comprising the steps:

- Providing a composition of the invention ("slurry"). Optionally, particulate graphite or conductive carbon black may be added as a conductive material to the mixture. - Apply the mixture to a current conductor.

- At least partial removal of liquid substances from the previously applied mixture.

The preparation of the electrode is initially carried out by coating the Stromab conductor. This can be realized by the process of casting, knife coating or printing the slurry onto the electrode conductor, followed by a drying step and subsequent calendering. The calendering is carried out in a manner that ensures the highest possible density of the electrode material, while maintaining a good pore structure, to ensure effective ion diffusion during the charging and discharging process. The electrode material layer should preferably also be characterized by good adhesion of the coating to the current conductor. As already mentioned, aluminum is preferred for the positive electrode and copper for the negative electrode in the conductors.

The present invention further provides an electrode obtainable by a method according to the invention and an electrochemical element comprising an electrode according to the invention, wherein the element is preferably a battery or an accumulator. The present invention is further illustrated by the following examples and figures, but is not limited thereto.

FIG. 1a shows the particle size distribution for a dispersion according to the invention

FIG. 1b shows the viscosity of a dispersion according to the invention

FIG. Figure lc shows a transmission electron micrograph of a device according to the invention

dispersion

FIG. ld shows the particle size distribution for a dispersion not according to the invention

FIG. 2 shows the particle size distribution for a dispersion according to the invention

FIG. 3a shows the particle size distribution for a dispersion according to the invention

FIG. 3b shows the viscosity of a dispersion according to the invention

FIG. 4a shows the particle size distribution for a dispersion according to the invention

FIG. 4b shows the viscosity of a dispersion according to the invention

FIG. 5 shows the specific electrical conductivity of electroplated CNT dispersions according to the invention, non-inventive CNT dispersions and conductive carbon black as conductivity additive FIG. 6 shows the results of adhesion tests of electrodes according to the invention and not according to the invention.

FIG. 7 shows different loading densities of the invention and not

electrodes according to the invention.

FIG. Figure 8 shows discharge capacities in consecutive cycles with batteries made with electrodes according to the invention and not according to the invention.

FIG. 9 shows the normalized specific discharge capacity for various

Discharge currents in batteries made with inventive and non-inventive electrodes.

FIG. 10 shows a SEM image of an electrode material according to the invention

FIG. 11 shows a SEM image of a non-inventive electrode material

FIG. 12a, 12b show further SEM images of the surface of an inventive

electrode material

FIG. Fig. 13 shows the particle size distribution for a comparative example

FIG. Fig. 14 shows the particle size distribution for another comparative example

FIG. Figure 15a shows the particle size distribution for a dispersion according to the invention

FIG. 15b shows the viscosity of a dispersion according to the invention

FIG. Fig. 15c shows a transmission electron micrograph of a device according to the invention

dispersion

FIG. 16 shows the particle size distribution for a dispersion according to the invention

FIG. 17 shows the particle size distribution for a dispersion according to the invention

FIG. Figure 18 shows the particle size distribution for dispersions of the invention

FIG. 19 shows the particle size distribution for dispersions according to the invention

The abbreviation "NMC" is used for the electrode active material LiNio, 33Mno, 33Coo, 3302 (Toda Kogyo Corp. Japan). PVDF stands for polyvinylidene fluoride (PVDF, SOLEF® 5130/1001, Solvay), PVP stands for polyvinylpyrrolidone (PVP K30, Sigma-Aldrich 81420), NMP stands for 1-methyl-2-pyrrolidinone (Sigma-Aldrich 328634), CMC stands for Carboxymethylcellulose (Walocel CRT 30G, Dow Chemicals). "Super P ^® Li" is a commercially available conductive carbon black (TIMCAL Graphite & Carbon, Switzerland). Unless otherwise indicated, CNTs used were Baytubes C70P (Bayer MaterialScience, Leverkusen) having a bulk density of about 70 g / dm ³ and a BET specific surface area of about 130 m ² / g or Baytubes C150P (Bayer MaterialScience, Leverkusen ) with a bulk density of about 150 g / dm ³ and with a BET specific surface area of about 210 m ² / g. Example la: Preparation of a dispersion according to the invention with NMP as dispersion medium 50 g of carbon nanotubes type C70P (Bayer MaterialScience, Leverkusen) with a BET specific surface area of about 130 m ² / g were treated with a knife mill (Retsch, Grindomix GM300) for 60 min. ground. 5 g of polyvinylpyrrolidone (PVP K30, Sigma-Aldrich 81420) were completely dissolved in 945 g of 1-methyl-2-pyrrolidinone (Sigma-Aldrich 328634) with stirring. The millbase was then mixed with the prepared solution and 90 min. homogenized with a rotor-stator system (Fluid Kotthoff GmbH). Thereafter, the mass was placed in a reservoir equipped with a stirrer and from which this mass was fed to a jet disperser. The jet disperser was equipped with a circular nozzle with a diameter of 0.5 mm. A pump delivered the dispersion through the nozzle orifice at a pressure of 160 bar and then back into the reservoir. After 60 passages, a nozzle with a diameter of 0.4 mm was used, the pressure was increased to 190 bar and the dispersing process was carried out for a further 60 passages. The total energy input based on the CNT mass used was about 42,000 kJ / kg.

Analytical Results: FIG. 1a shows the particle size distribution for a dispersion obtained according to Example 1a. The data were obtained by laser diffraction on a Malvern Mastersizer MS2000 Hydro MU instrument. In FIG. 1a, the cumulative volume fraction Q3 is plotted against the equivalent particle size. When determining the particle size of carbon nanotubes by means of laser diffraction, it has to be taken into account that due to the narrow, elongated shape of the CNTs only an equivalent particle size for an assumed spherical morphology can be obtained. The volume-related proportion of particles having a size of <1 μιη is almost 100%.

FIG. 1b shows the results of a viscosity measurement for a dispersion obtained according to Example 1a. The data were recorded using a Rheometer from Anton Paar (MCR series). One recognizes the pseudoplastic behavior of the dispersion in a readily processable viscosity range according to the invention.

FIG. 1c shows a transmission electron micrograph of a dispersion prepared according to Example 1a. The white bar in the lower left corner of the shot represents the scale of 1 μιη again. One recognizes the high degree of separation and the high proportion of less fragmented CNTs. From FIG. It can be seen qualitatively directly that> 70% by weight of the CNTs have a length of more than 200 nm, which is an important prerequisite for the formation of a homogeneous CNT network in the electrode.

Example 1b (comparative example): 50 g of carbon nanotubes of type C70P (Bayer MaterialScience, Leverkusen) with a BET specific surface area of about 130 m ² / g were treated with a knife mill (Retsch, Grindomix GM300) for 90 min. ground. 5 g of polyvinylpyrrolidone (PVP K30, Sigma-Aldrich 81420) were completely dissolved in 945 g of 1-methyl-2-pyrrolidinone (Sigma-Aldrich 328634) with stirring. The millbase was then mixed with the prepared solution and 90 min. homogenized with a rotor-stator system (Fluid Kotthoff GmbH). This dispersion was not treated with an HPD system and thus is not according to the invention.

Analytical results:

FIG. ld shows the particle size distribution for a dispersion obtained according to Example 1b. The data were obtained by laser diffraction on a Malvern Mastersizer MS2000 Hydro MU instrument. In FIG. ld, the cumulative volume fraction Qj is plotted against the equivalent particle size. When determining the particle size of carbon nanotubes by means of laser diffraction, it has to be taken into account that due to the narrow, elongated shape of the CNTs only an equivalent particle size for an assumed spherical morphology can be obtained. The volume-related proportion of particles with a size of <1 μιη, is almost 0% or almost nonexistent. Thus, the dispersion prepared in Comparative Example 1b is not according to the invention.

Example 2: Preparation of a dispersion according to the invention with NMP as dispersion medium

6 g of carbon nanotubes type C150P (Bayer MaterialScience, Leverkusen) with a higher bulk density of about 150 g / dm ³ compared to the type C70P and with a BET specific surface area of about 210 m ² / g were measured using a knife mill ( Retsch, Grindomix GM300) 60 min. ground. 1.2 g of polyvinylpyrrolidone (PVP K30) (Sigma-Aldrich 81420) was completely dissolved in 192.8 g of 1-methyl-2-pyrrolidinone (Sigma-Aldrich 328634) with stirring. The millbase was then mixed with the prepared solution and 90 min. Homogenized with a rotor-stator system (Fluid Kotthoff GmbH). Thereafter, the mass was dispersed with the Micron LAB 40 batch homogenizer (APV Gaulin Deutschland GmbH, Lübeck, Germany) at a pressure of 1000 bar. The dispersion process was repeated twice. The total energy input based on the CNT mass used was about 10,000 kJ / kg. Analytical results:

FIG. 2 shows the particle size distribution for a dispersion obtained according to Example 2. The data were obtained by laser diffraction on a Malvern Mastersizer MS2000 Hydro MU instrument. In FIG. 2, the cumulative volume fraction Q3 is plotted against the equivalent particle size. The volume-related proportion of particles with a size of <1 μηι is about 89%. When determining the particle size of carbon nanotubes by means of laser diffraction, it has to be taken into account that due to the narrow, elongated shape of the CNTs only an equivalent particle size for an assumed spherical morphology can be obtained. Example 3: Preparation of a dispersion according to the invention with 10% solids content and with water as dispersion medium

20 g of carbon nanotubes of type C70P (Bayer MaterialScience, Leverkusen) with a BET specific surface area of about 130 m ² / g were removed with a knife mill (Retsch, Grindomix GM300) for 60 min. ground. 2 g of carboxymethylcellulose (CMC, CRT30G, Dow Chemicals) were completely dissolved in 178 g of water with stirring. The millbase was then mixed with the prepared solution and 90 min. homogenized with a rotor-stator system (Fluid Kotthoff GmbH). Thereafter, the mass was dispersed with the Micron LAB 40 batch homogenizer (APV Gaulin Deutschland GmbH, Lübeck, Germany) at a pressure of 1000 bar. The dispersing process was repeated 11 times. The total energy input based on the CNT mass used was about 12,000 kJ / kg.

Analytical results:

FIG. 3a shows the particle size distribution for a dispersion obtained according to example 3. The data were obtained by laser diffraction on a Malvern Mastersizer MS2000 Hydro MU instrument. In FIG. 3a, the cumulative volume fraction Qj is plotted against the equivalent particle size. The volume-related proportion of particles with a size of <1 μιη is about 87%. When determining the particle size of carbon nanotubes by means of laser diffraction, it has to be taken into account that due to the narrow, elongated shape of the CNTs only an equivalent particle size for an assumed spherical morphology can be obtained. FIG. 3b shows the viscosity for a dispersion obtained according to example 3. The data were recorded using a Rheometer from Anton Paar (MCR series). One recognizes the pseudoplastic behavior of the dispersion in a high, but still easily processable viscosity range.

Example 4 Preparation of a dispersion according to the invention with 10% solids content, with water as dispersion medium and with a mixture of dispersing aids. 20 g of carbon nanotubes of type C70P (Bayer MaterialScience, Leverkusen) with a BET specific surface area of about 130 m ² / g were removed with a knife mill (Retsch, Grindomix GM300) for 60 min. ground. One gram of carboxymethyl cellulose (CMC, CRT30G, Dow Chemicals) and 1 gram of polyvinylpyrrolidone (PVP K30, Sigma-Aldrich 81420) were completely dissolved in 178 grams of water with stirring. The millbase was then prepared with the Solution mixed and 90 min. Homogenized with a rotor-stator system (Fluid Kotthoff GmbH). Thereafter, the mass was dispersed with the Micron LAB 40 batch homogenizer (APV Gaulin Deutschland GmbH, Lübeck, Germany) at a pressure of 1000 bar. The dispersing process was repeated 11 times. The total energy input based on the CNT mass used was about 12,000 kJ / kg.

Analytical results:

FIG. 4a shows the particle size distribution for a dispersion obtained according to Example 4. The data were obtained by laser diffraction on a Malvern Mastersizer MS2000 Hydro MU instrument. In FIG. 4a, the cumulative volume fraction Q3 is plotted against the equivalent particle size. The volume-related proportion of particles with a size of <1 μιη is about 95%. When determining the particle size of carbon nanotubes by means of laser diffraction, it has to be taken into account that due to the narrow, elongated shape of the CNTs only an equivalent particle size for an assumed spherical morphology can be obtained. It can be seen here that a mixture of dispersing aids allows a still slightly higher degree of separation of the CNTs in the dispersion. FIG. 4b shows the viscosity for a dispersion obtained according to Example 4. The data were recorded using a Rheometer from Anton Paar (MCR series). One recognizes the pseudoplastic behavior of the dispersion in a high, but still readily processable viscosity range according to the invention.

Example 5 Preparation of Slurries According to the Invention from a Dispersion According to Example 1 with Different Weight Shares of Carbon Nanotubes

3 g of polyvinylidene fluoride (PVDF, SOLEF® 5130/1001, Solvay) were dissolved in about 25 ml of 1-methyl-2-pyrrolidinone (NMP, Sigma-Aldrich 328634) by stirring for 4 h (about 500 rpm) at 30 ° C solved. In this NMP / PVDF solution, 30 g of the dispersion from Example 1, which contained an amount of 1, 5 g CNT, and stirred for about 2.5 h at room temperature (about 2000 U / min.). Thereafter, 44.5 g of NMC active material (NM 3100, Toda Kogyo Corp.) and 1.0 g of graphite (KS6L, Timcal Co.) were added and heated at 700 rpm. 60 min. further stirred. The solids content of this slurry was 50 g and the solids content contained 6% by weight of PVDF, 3% by weight of CNT, 0.3% by weight of polyvinylpyrrolidone, 2% by weight of graphite and 88.7% by weight of NMC. active material. By varying the amount of dispersion added from Example 1, the amount of CNT in the slurry can be increased or decreased over a wide range. In a corresponding manner, in each case the added amount of NMC active material must be adjusted so that the total solids content is again 50 g. In this way, with a constant amount of PVDF and graphite, the CNT content can be varied over a wide range.

Example 6a (comparative example): Preparation of noninventive slurry with different proportions of conductive additive consisting of conductive carbon black 3 g of polyvinylidene fluoride (PVDF, SOLEF® 5130/1001, Solvay) were dissolved in about 50 ml of 1-methyl-2-pyrrolidinone (NMP, Sigma-Aldrich 328634) by stirring for 4 h (about 500 rpm) at 30 ° C solved. 3 g of carbon black (SuperP Li, Timcal) were added to this NMP / PVDF solution and stirred at room temperature for about 2.5 h (about 2000 rpm). Thereafter, 43 g of NMC active material (NM 3100, Toda Kogyo Corp.) and 1.0 g of graphite (KS6L, Timcal Co.) were added and heated at 700 rpm. 60 min. further stirred. The solids content of this slurry was 50 g and the solids content contained 6% by weight of PVDF, 6% by weight conductive black, 2% by weight graphite and 86% by weight NMC active material. By varying the amount of conductive carbon black added, the amount of conductive additive in the slurry can be increased or decreased over a wide range. In a corresponding manner, in each case the added amount of NMC active material must be adjusted so that the total solids content is again 50 g. In this way, with a constant amount of PVDF and graphite Leitruß the proportion can be varied over a wide range.

Example 6b (Comparative Example) Preparation of Noninventive Slurry Having Different Proportions of Carbon Nanotubes 3 g of polyvinylidene fluoride (PVDF, SOLEF® 5130/1001, Solvay) were dissolved in about 50 ml of 1-methyl-2-pyrrolidinone (NMP, Sigma-Aldrich 328634 ) by stirring for 4 h (about 500 rpm) at 30 ° C. In this NMP / PVDF solution, 3 g of carbon nanotubes (Baytubes C70P, Bayer Material Science), which previously 60 min with a knife mill (Retsch, Grindomix GM300) 60 min. were added, and stirred for about 2.5 h at room temperature (about 2000 U / min.). The dispersion containing carbon nanotubes was not treated with a high-pressure homogenizer and CNTs did not exhibit the size properties according to the invention. Thereafter, 43 g of NMC active material (NM 3100, Toda Kogyo Corp.) and 1.0 g of graphite (KS6L, Timcal Co.) were added and heated at 700 rpm. 60 min. further stirred. The solids content of this slurry was 50 g and the solids content contained 6 wt.% PVDF, 6 wt.% Conductive black, 2 wt.% Graphite and 86 wt.% NMC active material. By varying the amount of carbon nanotube added, the amount of conductivity additive in the slurry can be increased or decreased over a wide range. In a corresponding manner, in each case the added amount of NMC active material must be adjusted so that the total solids content is again 50 g. In this way, with a constant amount of PVDF and graphite, the carbon nanotube content can be varied over a wide range.

Example 7: Production of electrodes according to the invention with different proportions of conductive additive

If necessary, the slurry of Example 5 was diluted with 1-methyl-2-pyrrolidinone (NMP, Sigma-Aldrich 328634) to a viscosity between about 5 and about 30 Pa-s at a shear rate of 1 / s ( measured with rheometer from the company Anton Paar, MCR series). Subsequently, the slurry with a doctor (target value for the wet film thickness: 120 μιη) to a 30 μηι thick aluminum foil aufgerakelt. This film was then dried at 60 ° C for 18 hours. Subsequently, this dried film was pressed (calendered) at a pressure of 7000 kg / cm ² .

Example 8 Production of Noninventive Electrodes Having Different Proportions of Conductive Additive

If necessary, the slurries of Examples 6a and 6b (Comparative Examples) were diluted with 1-methyl-2-pyrrolidinone (NMP, Sigma-Aldrich 328634) to a viscosity between about 5 and about 30 Pa.s at a shear rate of 1 / s (measured with rheometer from the company Anton Paar, MCR series). Subsequently, the slurries were knife-coated with a doctor blade (nominal value for the wet film thickness: 120 μm) onto a 30 μm thick aluminum foil. These films were then dried at 60 ° C for 18 hours. Subsequently, these dried films were pressed (calendered) at a pressure of 7000 kg / cm ² .

Example 9: Conductivities of the electrodes

To measure the conductivities, the inventive slurry (Example 5) suitable for producing an electrode (Example 7) was applied to a glass sheet by means of a doctor blade (target value for the wet film thickness: 120 μm) and dried at 60 ° C. for 18 hours. Various films with different levels of Carb on Nanotub were used from the dispersions of the invention, so that the dried films of 6 wt .-% PVDF, 2 wt .-% graphite and 1, 2, 3, 4, 6, 8 or 12 wt .-% carbon nanotubes from the dispersions of the invention passed. The respective difference to 100 wt .-% consisted of polyvinylpyrrolidone (each 1/10 of the proportion of carbon nanotubes) and active material NMC. Subsequently, the resistivities were measured by the 4-point method known to those skilled in the art.

For comparison, electrode films were prepared under the same conditions, but instead of slurries containing carbon nanotubes according to the invention (Example 5) slurries according to the invention containing conductive black (SuperP Li, Timcal) (Example 6a) or the carbon nanotubes from the non-inventive dispersions (Example 6b ) used.

Analytical results:

FIG. Figure 5 shows the specific conductivity of films prepared according to Example 9. The films consisted of 6 wt .-% PVDF, 2 wt .-% graphite and 1, 2, 3, 4, 6, 8 or 12 wt .-% carbon nanotube from the dispersion of the invention (solid line) or not Carbon nanotube-containing dispersion according to the invention (short-dashed line) or Leitruß from the dispersion not according to the invention (long dashed line). Clearly, the over many areas is to see many times greater specific conductivity of the electrode materials, if the dispersion according to the invention containing carbon nanotubes or the slurry prepared therefrom according to Example 5 is used instead of conductive carbon black or instead of carbon nanotubes not dispersed according to the invention. The higher conductivity is an important prerequisite for a higher power density of an electrochemical element (eg battery), in whose preparation the dispersion according to the invention was used.

Example 10: Adhesion Tests

Electrodes as described in Examples 7 and 8 were provided with a 10 mm wide adhesive strip on the electrode and pulled off with a tensile testing machine. The tensile force and thus the adhesion to the substrate were measured. The measurement was carried out in accordance with DIN EN ISO 11339. The results are shown in FIG. 6, which sets the force F in relation to the distance d over which the adhesive strip is pulled from the substrate. In this case, the upper curve 10 reproduces the measured values for the electrode material according to the invention and the lower curve 20 the measured values for the comparison material. It clearly shows the higher adhesion of the electrode according to the invention compared to the reference electrode. Example 11: load density

Electrode materials were prepared in a similar procedure as described in Examples 7 and 8. The composition of the electrode of the present invention was 89.8% by weight of NMC active material (NM3100, Toda Kogyo Corp.), 6.9% by weight of PVDF (Solef 5130, Solvay), 0.3% by weight of polyvinylpyrrolidone and 3 Wt .-% carbon nanotubes from the dispersion of the invention. The composition of the noninventive electrode was 89.8% by weight of NMC active material (NM3100, Toda Kogyo Corp.), 6.9% by weight of PVDF (Solef 5130, Solvay) and 3.3% by weight of conductive black (US Pat. SuperP Li, Timcal). The electrode films were pressed together under different pressures (calendered) and the density σ was determined by determining the mass of the electrode layer and the layer thickness of the test specimens obtained. The results are shown in FIG. 7 is shown. The square data points "■" relate to the material according to the invention and the triangular data points "A" to the comparison material. One can clearly see the higher density of the electrode according to the invention, which can be an important indicator of a higher energy density.

Example 12: Cyclization In this example, electrodes according to the invention and not according to the invention, as described in Examples 7 and 8 (in the case of non-inventive Example 8, based on the dispersion of Example 6a), were investigated with regard to their behavior during repeated charging and discharging. For this purpose, a model system in the form of a coin cell with a lithium anode and a cathode of LiNio, 33Mn ₀ , 33Coo, 330 ₂ (NMC) (NM3100, Toda Kogyo Corp.). The The composition of the cathode according to the invention was 85.7% by weight of NMC, 6% by weight of CNT from the dispersion according to the invention, 2% by weight of graphite (SG6L, Timcal), 0.3% by weight of polyvinylpyrrolidone (K30, Aldrich) and 6 Weight% PVDF binder. The composition of the comparison cathode not according to the invention was 86% by weight of> NMC, 3% by weight of SuperP Li, 2% by weight of graphite (SG6L, Timcal), and 7.2% by weight of PVDF binder. Thus, cells were available which contain the electrodes according to the invention and the reference electrodes but are otherwise identical. The particle size of the NMC particles was 5-10 μιη, the layer thickness of the electrode 60-70 μιη and the density of the electrode about 2.8 g / cm ³ . The electrolyte used was LP 30 Selectipur from Merck KGaA (1 M LiPF 6 in a 1: 1 ethylene carbonate / dimethyl carbonate (EC / DMC) mixture).

FIG. 8 shows the discharge capacity as a function of the number of charge / discharge cycles n. The charge and discharge current is C / 5, i. the full capacity was evenly loaded and unloaded over a period of 5h. After 100 cycles, a charge and discharge current of C was set for 10 cycles, i. the full capacity was then evenly charged and discharged over a period of 1 h. Although this procedure shows a significant decline in capacity at high discharge currents, but also a good regenerability of both systems after returning to low discharge currents. Data points with triangles "A" pointing upwards belong to electrodes according to the invention and those with triangles "T" facing down to the comparative example. Clearly recognizable is the much higher cycle life of a battery containing CNTs as a conductivity additive compared to the reference battery based on Leitruß as an additive. This example suggests a longer lifetime of a CNT-based battery.

Example 13: Load Capacity

Inventive and not inventive electrodes, as described in principle in Examples 7 and 8 were, as described in principle in Example 12, processed into button cells. Only the composition was changed as follows. The composition of the cathode according to the invention was 89.5% by weight of> NMC, 3% by weight of> CNT from the dispersion according to the invention, 0.3% of> polyvinylpyrrolidone and 7.2% by weight of PVDF binder. The composition of the comparison cathode not according to the invention was 89.8% by weight NMC, 3% by weight SuperP Li and 7.2% by weight PVDF binder. Thus, there were available cases which contained the electrodes according to the invention and the comparative electrodes but are otherwise identical in construction. The cells were then exposed to charging and discharging currents ranging from C / 5 (one full capacity in 5 hours) to 10C (10 full capacities in 1 hour). Each data point results from the mean value of five consecutive individual measurements with the same discharge rate. FIG. 9 shows the normalized specific discharge capacity for different discharge currents from C / 5 to I OC and the capacity that is present after the maximum nor at the minimum discharge current. The diamond-shaped data points "♦" belong to the electrode according to the invention and square data points "■" to the comparative example. It can be seen clearly that under the same conditions, the decrease of the capacity is lower with higher discharge currents, which speaks for a lower internal resistance of this battery and thus enables a higher power density. In addition, it can be seen that the battery with the non-inventive electrode has suffered a significant loss of capacity (about 30%) after treatment with the maximum discharge current, while the battery with the electrode according to the invention almost reaches the starting value again. This behavior showed the significantly improved performance of a battery containing the electrode according to the invention when subjected to a high load.

FIG. 10 shows a scanning electron micrograph of the cross-section of a refracted electrode according to the invention (not pressed). It was obtained on the basis of Example 7, the composition being 89.8% by weight of NMC, 2% by weight of CNTs, 2% by weight of graphite, 6% by weight of PVDF binder and 0.2% by weight of> polyvinylpyrrolidone. It can be seen a dense network of CNTs 30, which cover the NMC particles 40 without agglomerate-like accumulations of CNTs can be seen. This optimal distribution of CNTs ensures effective, low-resistance dissipation of the electrons from the active material to the metallic arrester. At the same time, this elastic CNT network ensures that expansions and contractions of the active material during charging and discharging do not result in loss of electrical contacts from the active material to the arrester.

FIG. 11 shows a scanning electron micrograph of the cross section of a noninventive electrode (not pressed). It was obtained according to Example 8 (involving the slurry of Example 6b), the composition being 90% by weight of> NMC, 2% by weight of CNTs, 2% by weight of graphite and 6% by weight of PVDF binder. The used CNTs are agglomerated here. One recognizes the agglomerates 50, the NMC particles 60 and graphite particles 70. The CNT agglomerates concentrate a larger part of the existing CNTs in narrowly defined areas, so that the overall conductivity of the electrode is inferior. Although this effect is also present at high CNT concentrations, it becomes all the more important with decreasing CNT concentrations in the electrode, since then increasingly percolation paths are interrupted. This also explains the behavior in FIG. 5, where it is clearly visible how the conductivity decreases at low concentrations. At low concentrations, the electrodes made of the noninventive dispersions behave similarly to electrodes containing conductive black as an additive. It does not form the active material surrounding network. Accordingly, the lower Zyklisierbarkeit, as shown in FIG. 8 is shown.

FIG. 12a and 12b show two scanning electron micrographs of the surface of an electrode according to the invention (not pressed). The composition of the electrode was 3% by weight of CNT from the dispersion according to the invention, 0.3% by weight of polyvinylpyrrolidone (K30, Aldrich) and 6% by weight of PVDF binder.

Example 14 (Comparative Example): Repetition of Example 3 from DE 10 2005 043 054 A1 (WO 2007/028369 A1)

10 g of carbon nanotubes of the type Baytubes C150P (Bayer MaterialScience, Leverkusen) with a BET specific surface area of about 210 m ² / g, an outer diameter of 13-16 nm and lengths of the individual carbon nanotubes of 1 to 10 μm were added without addition of further additives dispersed in 500 ml of 2-propanol, so that a dispersion with 1, 96% solids content was formed. Thereafter, the mass was dispersed with the batch homogenizer Micron LAB 40 (APV Gaulin Germany GmbH, Lübeck, Germany) at a pressure of 1000 bar with a single pass. That in DE 10 2005 043 054 Al (WO 2007/028369 Al) only one dispersion pass was passed, it follows from the formulation "after the passage was the dispersion ..." in the description of the example.

Analytical results:

FIG. Figure 13 shows the particle size distribution for a dispersion obtained according to Example 14. The data were obtained by laser diffraction on a Malvern Mastersizer MS2000 Hydro MU instrument. In FIG. 13, the cumulative volume fraction Qj is plotted against the equivalent particle size. The volume-related proportion of particles with a size of <1 μιη is almost nonexistent and is in the range of less than 2% and thus well below the limits according to the invention. The average agglomerate size (d50 value) is about 22 μm and the d90 value is more than 40 μm. When determining the particle size of carbon nanotubes by means of laser diffraction, it has to be taken into account that due to the narrow, elongated shape of the CNTs only an equivalent particle size for an assumed spherical morphology can be obtained.

Example 15 (Comparative Example): Preparation of a dispersion not according to the invention with N-methylpyrrolidone as dispersion medium

1 g of polyvinylpyrrolidone (PVP K30, Sigma-Aldrich 81420) was completely dissolved in 499 g of 1-methyl-2-pyrrolidinone (Sigma-Aldrich 328634) with stirring. Subsequently, 10 g of Baytubes C150P carbon nanotubes (Bayer MaterialScience, Leverkusen) with a BET specific surface area of about 210 m ² / g, an outer diameter of 13-16 nm and lengths of the individual carbon nanotubes of 1 to 10 μηι added to the solution, so that a dispersion with 1.96% CNT solids content was formed. Thereafter, the mass was dispersed with the batch homogenizer Micron LAB 40 (APV Gaulin Germany GmbH, Lübeck, Germany) at a pressure of 1000 bar with a single pass. Analytical results:

FIG. Fig. 14 shows the particle size distribution for a dispersion obtained in Example 15. The data were obtained by laser diffraction on a Malvern Mastersizer MS2000 Hydro MU instrument. In FIG. 14, the cumulative volume fraction Q3 is plotted against the equivalent particle size. The volume-related proportion of particles with a size of <1 μιη is about 20% and thus well below the limits of the invention. The size distribution is also very broad with a d50 value of about 7 μιη and a d90 value of almost 70 μιη. When determining the particle size of carbon nanotubes by means of laser diffraction, it has to be taken into account that due to the narrow, elongated shape of the CNTs only an equivalent particle size for an assumed spherical morphology can be obtained.

Example 16: Preparation of a Dispersion According to the Invention with NMP as Dispersion Medium and Ethyl Cellulose as Dispersing Aid

6 g of carbon nanotubes of type C70P (Bayer MaterialScience, Leverkusen) with a BET specific surface area of about 130 m ² / g were treated with a knife mill (Retsch, Grindomix GM300) for 60 min. ground. 0.3 g of ethylcellulose (EC, ETHOCELL Standard 100, Dow Wolff Cellulosics) were completely dissolved in 193.7 g of NMP with stirring. The millbase was then mixed with the prepared solution and 90 min. homogenized with a rotor-stator system (Fluid Kotthoff GmbH). Thereafter, the mass was dispersed with the Micron LAB 40 batch homogenizer (APV Gaulin Deutschland GmbH, Lübeck, Germany) at a pressure of 1000 bar. The dispersing process was repeated 5 times.

Analytical results:

FIG. Figure 15a shows the particle size distribution for a dispersion obtained according to Example 16. The data were obtained by laser diffraction on a Malvern Mastersizer MS2000 Hydro MU instrument. In FIG. 15a, the cumulative volume fraction Q3 is plotted against the equivalent particle size. The volume-related proportion of particles with a size of <1 μιη is about 95%. When determining the particle size of carbon nanotubes by means of laser diffraction, it has to be taken into account that due to the narrow, elongated shape of the CNTs only an equivalent particle size for an assumed spherical morphology can be obtained. FIG. Figure 15b shows the viscosity for a dispersion obtained according to Example 16. The data were recorded using a Rheometer from Anton Paar (MCR series). One recognizes the pseudoplastic behavior of the dispersion in a readily processable viscosity range.

FIG. Figure 15c shows a transmission electron micrograph of a dispersion prepared according to Example 15. The white bar in the lower right corner of the picture shows the scale of 500 nm. One recognizes the high degree of separation and the high proportion of less fragmented CNTs. From FIG. It can be seen qualitatively directly that also EC acts as an effective dispersing assistant and that> 70% by weight of the CNTs have a length of more than 200 nm, which is an important prerequisite for the formation of a homogeneous CNT network in the electrode.

Example 17: Preparation of a Dispersion According to the Invention with NMP as Dispersion Medium and EC as Dispersing Aid

40 g of carbon nanotubes of type C70P (Bayer MaterialScience, Leverkusen) with a BET specific surface area of about 130 m ² / g were removed with a knife mill (Retsch, Grindomix GM300) for 60 min. ground. 4 g of ethylcellulose (EC, ETHOCELL Standard 100, Dow Wolff Cellulosics) were completely dissolved in 956 g of 1-methyl-2-pyrrolidinone (Sigma-Aldrich 328634) with stirring. The millbase was then mixed with the prepared solution and 90 min. homogenized with a rotor-stator system (Fluid Kotthoff GmbH). Thereafter, the mass was placed in a reservoir equipped with a stirrer and from which this mass was fed to a jet disperser. The jet disperser was equipped with a circular nozzle with a diameter of 0.5 mm. A pump delivered the dispersion through the nozzle orifice at a pressure of 250 bar and then back into the reservoir. In total, 120 passages were driven.

Analytical Results: FIG. 16 shows the particle size distribution for a dispersion obtained according to Example 17. The data were obtained by laser diffraction on a Malvern Mastersizer MS2000 Hydro MU instrument. In FIG. 16, the cumulative volume fraction Qj is plotted against the equivalent particle size. When determining the particle size of carbon nanotubes by means of laser diffraction, it has to be taken into account that due to the narrow, elongated shape of the CNTs only an equivalent particle size for an assumed spherical morphology can be obtained. The volume-related proportion of particles having a size of <1 μιη, is about 93%.

Example 18 Preparation of a Dispersion According to the Invention with NMP as Dispersion Medium and Methylcellulose as Dispersing Aid 6 g of carbon nanotubes of type C70P (Bayer MaterialScience, Leverkusen) with a BET specific surface area of about 130 m ² / g were treated with a knife mill (Retsch, Grindomix GM300) for 60 min. ground. 0.3 g of methylcellulose (MC, Methocell, Sigma Aldrich) were completely dissolved in 193.7 g of NMP with stirring. The millbase was then mixed with the prepared solution and 90 min. homogenized with a rotor-stator system (Fluid Kotthoff GmbH). Thereafter, the mass was dispersed with the Micron LAB 40 batch homogenizer (APV Gaulin Deutschland GmbH, Lübeck, Germany) at a pressure of 1000 bar. The dispersing process was repeated 5 times.

Analytical Results: FIG. 17 shows the particle size distribution for a dispersion obtained according to Example 18. The data were obtained by laser diffraction on a Malvern Mastersizer MS2000 Hydro MU instrument. In FIG. 17, the cumulative volume fraction Qj is plotted against the equivalent particle size. The volume-related proportion of particles with a size of <1 μιη is about 93%. When determining the particle size of carbon nanotubes by means of laser diffraction, it has to be taken into account that due to the narrow, elongated shape of the CNTs only an equivalent particle size for an assumed spherical morphology can be obtained.

EXAMPLE 19 Preparation of a Dispersion According to the Invention with NMP as Dispersion Medium and Polyvinylpyridine as Dispersing Aid 6 g of carbon nanotubes of the type C70P (Bayer MaterialScience, Leverkusen) with a BET specific surface area of about 130 m ² / g were measured using a knife mill (Retsch, Grindomix GM300) 60 min. ground. 0.6 g of poly (4-vinylpyridine) (Mw = 60,000, Aldrich) was completely dissolved in 193.4 g of NMP with stirring. The millbase was then mixed with the prepared solution and 90 min. Homogenized with a rotor-stator system (Fluid Kotthoff GmbH). Thereafter, the mass was dispersed with the Micron LAB 40 batch homogenizer (APV Gaulin Deutschland GmbH, Lübeck, Germany) at a pressure of 1000 bar. The dispersing process was repeated 5 times.

EXAMPLE 20 Preparation of a Dispersion According to the Invention Using NMP as Dispersion Medium and Polystyrene as Dispersing Aid 6 g of carbon nanotubes of the type C70P (Bayer MaterialScience, Leverkusen) with a BET specific surface area of about 130 m ² / g were removed using a knife mill (Retsch, Grindomix GM300) 60 min. ground. 0.6 g of polystyrene (Mw = 60,000, Aldrich) was completely dissolved in 193.4 g of NMP with stirring. The millbase was then mixed with the prepared solution and 90 min. Homogenized with a rotor-stator system (Fluid Kotthoff GmbH). Thereafter, the mass was dispersed with the Micron LAB 40 batch homogenizer (APV Gaulin Deutschland GmbH, Lübeck, Germany) at a pressure of 1000 bar. The dispersing process was repeated 5 times.

Example 21 Preparation of a Dispersion According to the Invention with Water as Dispersion Medium and a Mixture of Polyacrylic Acid and CMC as Dispersing Aid

12 g of carbon nanotubes type C70P (Bayer MaterialScience, Leverkusen) with a BET specific surface area of about 130 m ² / g were treated with a knife mill (Retsch, Grindomix GM300) for 60 min. ground. 2.4 g of CMC (Mw = 60,000, Aldrich) were completely dissolved in 385.6 g of water with stirring. The millbase was then mixed with the prepared solution and 90 min. Homogenized with a rotor-stator system (Fluid Kotthoff GmbH). Thereafter, the mass was dispersed with the Micron LAB 40 batch homogenizer (APV Gaulin Deutschland GmbH, Lübeck, Germany) at a pressure of 1000 bar. The dispersing process was repeated 5 times. Subsequently, polyacrylic acid (PAA, Mw ~ 240,000, Sigma Aldrich) was adjusted to pH 8.5 with LiOH so that an aqueous 28.8% solution was available. 100 g of this solution were then mixed with 200 g of the carbon nanotube dispersion and dispersed once with the Micron LAB 40 batch homogenizer (APV Gaulin Deutschland GmbH, Lübeck, Germany) at a pressure of 1000 bar. This gives a stable, finely divided aqueous dispersion consisting of 2% by weight of carbon nanotubes, 9.6% polyacrylic acid and 0.4% CMC. For a comparative dispersion, a procedure was used according to Example 3, in which instead of CMC only PAA adjusted to pH 8.5 with LiOH was used.

Analytical results:

FIG. Fig. 18 shows the particle size distribution for a solid line dispersion obtained in Example 21). For comparison, the particle size distribution of a dispersion stabilized only with polyacrylic acid (dashed line). The data were obtained by laser diffraction on a Malvern Mastersizer MS2000 Hydro MU instrument. In FIG. 18, the cumulative volume fraction Qj is plotted against the equivalent particle size. The volume-related proportion of particles with a size of <1 μιη is about 95%. When determining the particle size of carbon nanotubes by means of laser diffraction, it has to be taken into account that due to the narrow, elongated shape of the CNTs only an equivalent particle size for an assumed spherical morphology can be obtained.

Example 22 Preparation of a Dispersion According to the Invention with a Carbon Nanotube with a High Specific Surface Area, NMP as Dispersion Medium and Ethyl Cellulose as Dispersing Aid 6 g of carbon nanotubes of the type Nanocyl NC 7000 (NANOCYL SA, Belgium) with a BET specific surface area of about 250-300 m ² / g were removed with a knife mill (Retsch, Grindomix GM300) for 60 min. ground. 2.4 g of ethylcellulose (EC, ETHOCELL Standard 100, Dow Wolff Cellulosics) were completely dissolved in 191.6 g of NMP with stirring. The millbase was then mixed with the prepared solution and 90 min. homogenized with a rotor-stator system (Fluid Kotthoff GmbH). Thereafter, the mass was dispersed with the Micron LAB 40 batch homogenizer (APV Gaulin Deutschland GmbH, Lübeck, Germany) at a pressure of 1000 bar. The dispersing process was repeated 5 times. In a further experiment, 6 g of carbon nanotubes are processed by the Practice AMC (UBE Industries, Japan), 1.2 g of ethylcellulose (EC, ETHOCELL Standard 100, Dow Wolff Cellulosics) together with 192.8 g of NMP in an analogous process.

Analytical results:

FIG. 19 shows the particle size distribution for the dispersions obtained according to Example 22. The data were obtained by laser diffraction on a Malvern Mastersizer MS2000 Hydro MU instrument. In FIG. 19, the cumulative volume fraction Qj versus the equivalent particle size is plotted for both dispersions (Nanocyl NC7000: solid line, AMC: dashed line). The volume-related proportion of particles with a size of <1 μm is about 87% in the dispersion produced with Nanocyl NC 7000. In the case of the dispersion produced with Üb AMC, this is almost 100%. When determining the particle size of carbon nanotubes by means of laser diffraction, it has to be taken into account that due to the narrow, elongated shape of the CNTs only an equivalent particle size for an assumed spherical morphology can be obtained.

This result demonstrates the universality of the process described in this invention, which is applicable not only to the Bayer MaterialScience C70P or C150P types, but also to types from other manufacturers.

Claims

claims

1. A dispersion comprising a dispersion medium, a polymeric dispersing aid and dispersed in the dispersion medium carbon nanotubes, characterized in that the proportion of present in agglomerates with an average agglomerate size of> 1 μιη carbon nanotubes in the total amount of carbon nanotubes <40% by volume and that > 70% by weight of the carbon nanotubes present in unagglomerated form have a length of> 200 nm.

2. Dispersion according to claim 1, wherein the dispersion medium is selected from the group consisting of water, acetone, nitriles, alcohols, dimethylformamide (DMF), N-methylpyrrolidone (NMP), pyrrolidone derivatives, butyl acetate, methoxypropyl acetate, alkylbenzenes, cyclohexane derivatives and mixtures thereof.

3. Dispersion according to claim 1 or 2, wherein the dispersing aid is selected from the group poly (vinylpyrrolidone) (PVP), polyvinylpyridines (eg poly (4-vinylpyridine) or poly (2-vinylpyridine)), polystyrene (PS), poly ( 4-vinylpyridine-co-styrene), poly (styrenesulfonate) (PSS), lignin sulfonic acid, lignosulfonate, poly (phenylacetylene) (PPA), poly (meta-phenylenevinylene) (PmPV), polypyrrole (PPy), poly (p-phenylenebenzobisoxazole) (PBO), naturally occurring polymers, anionic aliphatic surfactants, poly (vinyl alcohol) (PVA), polyoxyethylene surfactants, poly (vinylidene fluoride) (PVdF), cellulose derivatives, mixtures of various cellulose derivatives, polyvinyl chloride (PVC), polysaccharides, styrene-butadiene Rubber (SBR), polyamides, polyimides, block copolymers (for example, acrylic block copolymers, ethylene oxide-propylene oxide copolymers) and mixtures thereof.

4. Dispersion according to one or more of claims 1 to 3, wherein the dispersing aid comprises lithium ions.

5. Dispersion according to one or more of claims 1 to 4, wherein the ratio of the concentration of the dispersing aid in the dispersion medium and the concentration of

Carbon nanotubes in the dispersion medium in a range of> 0.01: 1 to <10: 1.

6. Dispersion according to one or more of claims 1 to 5, wherein dispersion further comprises Leitruß, graphite and / or graphene.

7. Process for the preparation of a dispersion according to one or more of claims 1 to 6, characterized in that a precursor dispersion comprising a dispersion medium, a poly mer s D isp ergi erhi l fs mitt el and Ko hle ns to ffn or else in high pressure homogenizer.

8. The method according to claim 7, wherein the dispersion is carried out several times by means of the high-pressure homogenizer.

9. The method of claim 7 or 8, wherein the high-pressure homogenizer is a jet disperser and at least one nozzle having a bore diameter of> 0.05 to <1 mm and a length to diameter ratio of the bore of> 1 to <10, wherein between Nozzle inlet and nozzle outlet a pressure difference of> 5 bar exists.

10. The method according to claim 9, wherein the jet disperser has at least one slot gap with a gap width of> 0.05 to <1 mm and a depth to gap width ratio of the slot gap of> 1 to <10, wherein between the nozzle inlet and nozzle outlet a pressure difference of > 5 bar.

1 1. A composition for producing an electrode comprising a dispersion according to one or more of claims 1 to 6, an electrode material and a polymeric binder, wherein the binder is present at least partially in dissolved form in the composition.

12. The composition according to claim 11, wherein the electrode material is selected from the group LiNi _x Mn _y Al _z COI _x - _y - _z C> 2 (0 <x, y, z <l and x + y + z < 1), LiNio, 33Mno, 33Coo, 330 ₂ , LiCoO ₂ ,

LiNi ₀ , 7Coo, 30 ₂ , LiNi ₀ , 8Coo, ₂ 0 ₂ , LiNio, 9Co ₀ , i0 ₂ , LiNi0 ₂ , LiMn ₂ 0 ₄ , LiMni. ₅ (Co, Fe, Cr) o, ₅ C> 4, LiNi _x Al _y Coi-x-y0 ₂ (0 <x, y <l and x + y <l), LiNi ₀ , 8Coo, i5Al ₀ , o50 ₂ , LiNi ₀ , 78Coo, i9Al ₀ , o30 ₂ , LiNio, 78Coo, i9Al ₀ , o3M _x 0 ₂ (x = 0.0001-0.05, M = alkali or alkaline earth metals), LiFeP0 ₄ , Li ₂ FeP ₂ 0 ₇ , L1C0PO4, Lii _{+ x} M _y Mn ₂ . _x . _y 0 ₄ (M = Al, Cr, Ga), LiTiS ₂ , Li ₂ V ₂ 0 ₅ , LiV ₃ 0 ₈ , Li ₂ TiS ₃ , Li ₃ NbSe ₃ , Li ₂ Ti0 ₃ , sulfur, polysulfides and / or sulfur-containing Materials.

A composition according to claim 11, wherein the electrode material is selected from the group consisting of natural or synthetic graphite, hard carbon, which has a stable random structure of highly crosslinked very small and thin carbon platelets, soft, graphitic carbon, silicon, silicon alloys, silicon containing mixtures, lithium titanate (Li ₂ TiO 3 or LLiTisOn), tin alloys, C03O4, Li ₂ , 6CoO, 4N and / or tin oxide (SnO ₂ ).

14. A method of making an electrode comprising the steps of:

- providing a composition according to claim 11, 12 or 13;

- applying the composition to a current conductor;

15. An electrochemical element comprising an electrode prepared according to claim 14.