DISPERSANT

Abstract

The present invention relates to dispersants comprising a polyester derived from a plurality of poly(carbonylalkyleneoxy) chains (and optionally a fatty acid) and an amine, wherein said dispersants have an acid value of less than 15 mgKOH/g. The dispersants as described herein provide desirable viscosity profiles when in use. More especially some embodiments provide improved compatibility with certain solvents in dispersions, and/or are particularly suitable for use in battery systems.

Claims

1. A dispersant comprising: a polyester derived from a plurality of poly(carbonylalkyleneoxy) chains and an amine, wherein the dispersant has an acid value of less than 5 mgKOH/g.

2. A dispersant according to claim 1, wherein the dispersant comprises a fatty acid selected from one or more of the following: oleic acid, caproic acid, lauric acid, stearic acid and palmitic acid.

3.-4. (canceled)

5. A dispersant according to claim 1, wherein the dispersant has an acid value of less than 1 mgKOH/g.

6. A dispersant according to claim 1 wherein the plurality of poly(carbonylalkyleneoxy) chains contains an alkylene group having between 3 and 12 carbon atoms.

7. A dispersant according to claim 6, wherein the alkylene group contains between 4 and 8 carbon atoms.

8. (canceled)

9. A dispersant according to claim 2, wherein a ratio of the plurality of poly(carbonylalkyleneoxy) chains to the saturated fatty acid is between 4:1 and 15:1.

10. A dispersant according to claim 1, wherein the amine is a poly(alkylenimine).

11. (canceled)

12. A dispersant according to claim 10, wherein the poly(alkylenimine) is polyethylenimine.

13. A dispersant according to claim 10, wherein the poly(alkylenimine) has a weight average molecular weight in the range of 1000 to 50000, preferably 2000 to 25000.

14. (canceled)

15. A dispersant according to claim 1, wherein the dispersant comprises at least 5 wt % (weight percent) amine and optionally no more than 35 wt % amine.

16.-18. (canceled)

19. A dispersant according to claim 1, wherein the dispersant has an amine value of less than 30 mgKOH/g.

20. (canceled)

21. A dispersant according to claim 1, wherein the dispersant has a weight loss of less than 20% at 350° C.

22. A method of manufacturing a dispersant in accordance with claim 1, the method comprising the following steps: a) preparing a polyester from the polymerisation reaction of a plurality of poly(carbonylalkyleneoxy) chains and optionally a fatty acid, b) providing an amine and allowing the polyester of step a) to react with the amine to form an intermediate reaction product, and, c) reducing the acid value of the intermediate reaction product of step b).

23. A method of manufacturing a dispersant according to claim 22, wherein the polymerisation reaction of step a) is performed in the presence of a polymerisation catalyst.

24. A method of manufacturing a dispersant according to claim 23, wherein the polymerisation catalyst is selected from the following: titanium (IV) butylate, zirconium (VI) butoxide, zinc acetate and toluenesulfonic acid.

25. A method of manufacturing a dispersant according to claim 22, wherein the polymerisation reaction of step a) is performed at a temperature of between 130° C. and 250° C.

26. (canceled)

27. A method of manufacturing a dispersant according to claim 22, wherein step b) has a reaction time of between 14 and 24 hours.

28. (canceled)

29. A method of manufacturing a dispersant according to claim 22, wherein in step c), the reduction in acid value of the intermediate reaction product, is achieved (or assisted) by the introduction of an acid scavenger.

30. A dispersion comprising a continuous phase, a particulate to be dispersed and a dispersant in accordance with claim 1.

31. A dispersion according to claim 30, wherein the continuous phase comprises a solvent.

32. A dispersion according to claim 31, wherein the solvent is N-methyl-2-pyrrolidone (NMP).

33. A dispersion according to claim 30, wherein the particulate to be dispersed is a particulate battery active material selected from one or more of the following: conductive carbon, lithium nickel manganese cobalt oxide (NMC, LiNi.sub.xMn.sub.yCo.sub.zO.sub.2), lithium manganese oxide (LMO, LiMn.sub.2O.sub.4), lithium iron phosphate (LFP, LiFePO.sub.4), lithium cobalt oxide (LCO, LiCOO.sub.2), and lithium nickel cobalt aluminium oxide (NCA, LiNiCoAlO.sub.2).

34.-35. (canceled)

36. A dispersion according to claim 33 comprising an amount of particulate to be dispersed of between 60 wt % and 99.9 wt % of the total dispersant.

37. A dispersion according to claim 30, further comprising an additional additive selected from one or more of binders, adhesion promoters, wetting agents, and corrosion inhibitors.

38. A battery system comprising a dispersion in accordance with claim 30.

39.-42. (canceled)

Description

[0071] The present invention will now be described with reference to the following examples and accompanying Figures in which,

[0072] FIG. 1. shows viscosity profile data expressed in centipoise for LITX 50 conductive carbon particulate dispersed in NMP solvent comparing as dispersant PVP, high acid value Hypermer KD1 (comparative), and low acid value sample 118 in accordance with the present invention.

[0073] FIG. 2. shows viscosity profile data expressed in centipoise for Super C65 conductive carbon particulate dispersed in NMP solvent comparing as dispersant PVP, high acid value Hypermer KD1 (comparative), and low acid value sample 118 in accordance with the present invention.

[0074] FIG. 3. shows viscosity profile data expressed in centipoise for a simple battery system cathode formulation comparing as dispersant PVP, high acid value Hypermer KD1 (comparative), and low acid value sample 118 in accordance with the present invention.

[0075] FIG. 4. shows concentration of dispersant data expressed in wt % (weight %) relative to dispersion system viscosity comparing as dispersant PVP, high acid value Hypermer KD1 (comparative), and low acid value sample 118 in accordance with the present invention.

[0076] FIG. 5. shows increased carbon concentration data expressed in wt % (weight %) relative to dispersion system viscosity comparing as dispersant PVP, high acid value Hypermer KD1 (comparative), and low acid value sample 118 in accordance with the present invention.

[0077] FIG. 6. shows particle size peak data for LITX 50 conductive carbon particulate dispersed in NMP solvent comparing as dispersant PVP, high acid value Hypermer KD1 (comparative), and low acid value dispersant sample 118 in accordance with the present invention.

[0078] FIG. 7. shows the discharge capacity as a function of C-rate for an LCO-based coin cell comparing a cathode slurry without a dispersant, with PVP as a dispersant, and with the low acid value dispersant sample 118 in accordance with the present invention.

[0079] FIG. 8. shows the discharge capacity as a function of C-rate for an NMC622-based coin cell comparing a cathode slurry without a dispersant, with PVP as dispersant, and with the low acid value dispersant sample 118 in accordance with the present invention.

EXAMPLES

[0080] In the following examples dispersants were prepared in accordance with the following preparation methods.

[0081] For each sample prepared the weight percent of the raw materials used was varied to provide a polyester having a ε-caprolactone : lauric acid ratio as indicated in Table 1 (where appropriate).

[0082] bi) Comparative Example—Hypermer KD1 ex. Croda.

[0083] Hypermer KD1 is a comparative commercially available material. A small scale process, in accordance with the teaching of published patent application number EP 0 158 406 as described below, was performed to provide a chemical equivalent to commercially available Hypermer KD1.

[0084] a) Lauric acid, ε-caprolactone and zirconium (IV) butoxide catalyst (used at a level of 0.1 wt %) were charged to a reaction vessel equipped with a nitrogen headspace purge and agitator. The reactants were mixed and heated up to and maintained at a temperature of 180° C. until the reaction was complete (GPC end point determination, <1 wt % of raw materials present in the GPC trace). The reaction product was then cooled to 100° C. and discharged from the reaction vessel.

[0085] This product was prepared from a polyester material as prepared as per step a) above (with reference to Table 1), and polyethylenimine which was charged to a reaction vessel equipped with a nitrogen headspace purge. The reactants were mixed and heated up to and maintained at 150° C. until the reaction was complete (acid value in the range of 25-30 mgKOH/g) after approximately 5 hours. The reaction was then cooled to 90° C. and discharged from the reaction vessel. It should be noted, however, that in the tests performed below a commercially available Hypermer KD1 product obtained from a plant batch was tested.

[0086] bii) Examples in accordance with the present invention—Method 1

[0087] (Table 1 Samples 120, 154A, 154B, 156A, 156B, 158, 159, and 118)

[0088] a) Lauric acid, ε-caprolactone and zirconium (IV) butoxide catalyst (used at a level of 0.1 wt %) were charged to a reaction vessel equipped with a nitrogen headspace purge and agitator. The reactants were mixed and heated up to and maintained at a temperature of 180° C. until the reaction was complete (GPC end point determination, <1 wt % of raw materials present in the GPC trace). The reaction product was then cooled to 100° C. and discharged from the reaction vessel.

[0089] The polyester materials as prepared in step a) above (with reference to Table 1), and polyethylenimine were charged to a reaction vessel equipped with a nitrogen headspace purge. The reactants were mixed and heated up to and maintained at 175° C. and the reaction was allowed to continue until the desired atypical low acid value was reached (the obtained acid values are indicated in Table 1). The reaction product was then cooled to 90° C. and discharged from the reaction vessel.

[0090] bii) Examples in accordance with the present invention — Method 2

[0091] (Table 1 Sample 197)

[0092] Sample 197 was prepared by an alternative process route, where step a) as described above was not utilised, instead this method concerns an alternative direct PEI initiated polycaprolactone process route.

[0093] ε-caprolactone, polyethylenimine and zirconium (IV) butoxide catalyst (used at a level of 0.1 wt %) were charged to a reaction vessel equipped with a nitrogen headspace purge, agitator and condenser set for condensate removal. The reactants were mixed and heated up to and maintained at a temperature of 180° C. until the reaction was complete, in this case this was approximately 3 hours (GPC end point determination was used to establish completion of reaction, <1 wt % of ε-caprolactone present in the GPC trace). The reaction product was then cooled to 100° C. and discharged from the reaction vessel.

[0094] For each sample prepared the weight percent of the raw materials used was varied to provide a dispersant containing a polyethylenimine weight percentage as indicated in Table 1.

[0095] The polyethylenimine (PEI) materials used are all commercially available, and their relevant weight average molecular weights (MW) are as indicated in Table 1.

[0096] Acid values were calculated in accordance with test method AOCS Te-1a-64: Acid Value.

TABLE-US-00001 TABLE 1 Ratio of ε- Sample caprolactone: PEI Molecular Acid Value Name lauric acid PEI wt % Weight (MW) mgKOH/g Hypermer KD1 10:1 7.0 10000 30.3 (Comparative) 120 05:1 7.5 10000 0.7 154A 08:1 7.5 10000 0.7 154B 12:1 7.5 10000 0.5 156A 10:1 5.0 10000 2.1 156B 10:1 10.0 10000 0.6 158 10:1 7.5 25000 0.5 159 10:1 7.5 2000 1.0 118 10:1 7.0 10000 0.6 197 Not 7.0 10000 1.5 Applicable

[0097] PEI Materials utilised are as follows:

[0098] 2,000 MW material is Lupasol PR8515 ex. BASF

[0099] 10,000 MW material is Epomin SP200 ex. Nippon Shokubai.

[0100] 25,000 MW material is Lupasol WF ex. BASF

[0101] Testing 1. Dispersion Viscosity

[0102] Samples of commercially available conductive carbon particulates were dispersed in NMP as solvent in the presence of a dispersant selected from either a benchmark commercially available PVP (ex. Sigma Aldrich molecular weight 20 000), the high acid value comparative sample detailed above (Hypermer KD1), or the low acid example sample 118 as described above.

[0103] Two commercially available conductive carbon particulates were tested:

[0104] 1) LITX 50 ex. Cabot which is a conductive additive utilised in lithium ion batteries for hybrid electric vehicles and high-end consumer electronic devices such as smartphones.

[0105] 2) Super C65 ex. Imerys Graphite & Carbon which is a conductive additive utilised in lithium ion rechargeable batteries.

[0106] An IKA overhead stirrer with a 4-blade propeller stirrer was used to prepare dispersions consisting of the materials detailed above, at ambient temperature. The dispersant was provided at a level of 1 wt. % and the carbon at a level of 5 wt. % of the total dispersion.

[0107] The viscosity profiles of the dispersions are shown in FIGS. 1 and 2 for LITX 50 and Super C65 respectively. The viscosity profiles were obtained using a TA Instruments DHR-2 rheometer fitted with a 40 mm stainless steel parallel plate with a gap of 500 micrometres. The low acid value sample in accordance with the present invention provides an improved lower viscosity profile compared to the high acid comparative sample, and even provides some improvements over the PVP benchmark sample when used in combination with Super C65.

[0108] In addition, a related simple battery system cathode formulation containing either the commercially available PVP (ex. Sigma Aldrich molecular weight 20 000) benchmark, the high acid value comparative sample, or the low acid sample 118 in accordance with the present invention as detailed above were tested, and the obtained viscosity profiles are shown in FIG. 3. These simple cathode formulations also contained LCO and LITX50 ex. Cabot as the active particulate materials to be dispersed, and the solvent in which they were dispersed was NMP.

[0109] In a battery system cathode formulation, it is important to have a low viscosity, coupled with a shear thinning behaviour; this allows the formulation to be easily deposited where required, but then to remain in place once deposited. In the present case the low acid value sample provides a lowered the viscosity profile compared the high acid value sample. This addresses a need identified by battery cell manufacturers who have reported undesirable formulation thickening when utilising the commercially available materials which accord with the high acid value sample. In addition, when compared to the PVP benchmark dispersant, the low acid value sample dispersant provides an improved shear thinning viscosity.

[0110] 2. Dispersion Performance

[0111] The performance of three dispersants, 1. PVP, 2. high acid value Hypermer KD1 (comparative), and 3. low acid value dispersant sample 118 in accordance with the present invention, to disperse LITX 50 conductive carbon in NMP solvent was considered in relation to concentration of dispersant required to achieve acceptable dispersion, loading level of particulate to be dispersed and particle size peak data of particulate dispersed. The results obtained are shown in FIGS. 4 to 6 and discussed below.

[0112] Dispersions were made by dissolving the dispersant in NMP by stirring with an overhead stirrer (IKA) equipped with a 4-blade impeller. LITX50 was added and stirring continued until homogeneous.

[0113] Viscosity measurements were carried out at 25 ° C. on a DHR-2 rheometer (TA Instruments) equipped with a 40 mm stainless steel parallel plate. The gap was set to 500 μm and a soak time of 30 seconds was allowed before the measurements were carried out. A flow curve was run logarithmically from a shear rate of 0.1 to 1000 s.sup.−1. Where specific data points have been used for comparison, these have been taken at 10 s.sup.−1.

[0114] FIG. 4 shows the amount of dispersant required to disperse LITX 50 conductive carbon in NMP solvent at an acceptable viscosity. In this case, it can be seen that only 2 wt % of example sample 118 was required to achieve an acceptable dispersion versus 6 wt % of Hypermer KD1. Furthermore, it can be seen that example sample 118 provides a dispersant concentration similar to that of PVP.

[0115] In addition, FIG. 5 shows that higher carbon loadings can be achieved with example sample 118 at a lower viscosity for all carbon loading levels tested as compared to Hypermer KD1. Furthermore, example sample 118 provides beneficial higher carbon loadings compared to PVP. As such, the materials of the present invention allow higher carbon loadings to be achieved for the same viscosity in the solvent system. This higher loading level capability is advantageous for use of the present dispersants in battery systems.

[0116] Particle size measurements were carried out on a CPS Disc Centrifuge (CPS Instruments) with 100 μl of particles measured from 1.0 to 0.01 μm at a disc speed of 20 000 rpm. The gradient used was comprised of NMP and Halocarbon 1.8.

[0117] FIG. 6 shows the particle size peak data obtained. The particle size peak data shows that the use of the 118 dispersant reduces the particle size as compared to use of PVP; this result is indicative of improved dispersion ability of example sample 118 over PVP. It can be seen that no improvement is provided over Hypermer KD1 (comparative example), however, the low acid value materials of the present invention avoid the viscosity increasing affects observed with use of Hypermer KD1, as discussed above, with no appreciable detriment to dispersion ability as compared to Hypermer KD1. As such, the present material offers benefits over both PVP and Hypermer KD1 which render them particularly suitable for use in battery systems, but also suitable for use in other dispersion systems where viscosity control and increased particulate loading may be attractive.

[0118] 3. Effect of Dispersant on Capacity in Lithium Ion Coin Cell Batteries

[0119] Two commercially available active battery materials were tested to assess the effect of the dispersant sample 118 on capacity of composite cathodes of lithium ion batteries:

[0120] 1) LCO ex. Gelon typically used in batteries for consumer electronics.

[0121] 2) NMC622 ex. Gelon which is a Lithium-Nickel-Manganese-Cobalt-Oxide (LiNiMnCoO2) active cathode composite powder typically used for automotive batteries.

[0122] The composition of the solid composite cathode materials prepared for the cathode capacity testing is indicated in Table 2, below. The carbon black utilised was commercially available CB grade LITX200 ex. Cabot. The binder utilised was a commercially available polyvinylidene fluoride (PVDF) powder Solef 6010/1001 ex. Solvay. The commercially available PVP dispersant was also tested as an alternative.

TABLE-US-00002 TABLE 2 Composite Cathode Material Content in wt. % Content, grams Active Cathode Battery 94.63 4.000 Material Carbon Black 1.99 0.084 Dispersant 0.40 0.017 Binder (PVDF) 2.98 0.126 Total 100.00 4.227

[0123] To prepare the cathode slurries for testing initially the carbon black was pre-dispersed in NMP solvent in a carbon: dispersant ratio 5:1 wt. % using a centrifugal mixer (a Thinky Mixer ARE-310). Subsequently the active cathode battery material (LCO or NMC622) and the PVDF binder were added to the initial carbon black dispersion and mixed for one hour in a SPEX—8000M ball mill. The final slurries contained 67 wt % solid composite cathode material loaded into 33 wt % NMP solvent.

[0124] A cathode slurry containing no dispersant (blank) was also prepared as a reference sample to be tested.

[0125] The cathode slurries to be tested where rendered as cathode layers in coin cells to be tested by providing a layer of the cathode slurry on battery-grade aluminium foil using a doctor blade applicator and subsequent drying at 110° C. in a vacuum oven. After drying, cathode discs were cut from the coated film using a high-precision disc puncher and the discs transferred into an argon-filled glovebox with an oxygen and moisture content of less than 1 ppm. 2032-type coin cells were assembled inside the glove box using the prepared cathode discs and lithium metal as a reference/counter electrode, porous polypropylene separator and commercial 1 M LiPF6 electrolyte solution in ethylene carbonate/diethyl carbonate 50/50 v/v. The prepared coin cells were left to rest for 12 hours prior to electrochemical testing. The coin cells were galvanostatically cycled five times in the potential ranges 3.2-4.2 V vs. Li/Li+ (LCO) and 3.0-4.3 V vs. Li/Li+ (NMC622) at currents corresponding to 0.1 C-rate for formation using a battery cycling setup. Then, cycling was carried out at 0.5 C, 1.0 C and 2.0 C (five charge-discharge cycles at each C-rate) using the same potential ranges. Capacity at each C-rate was calculated as an average over five cycles.

[0126] FIGS. 7 and 8 show the rate performance data (discharge capacity versus C-rate) of the coin cells containing LCO and NMC622 cathode active materials, respectively. Both graphs also show the results obtained for the blank samples (i.e. containing no dispersant) for comparison versus where PVP or Sample 118 are included as a dispersant. It can be seen that for both active battery material chemistries tested, using the carbon dispersant resulted in an increase of capacity; and the advantageous effect is greater at higher discharge rates. Improved capacity retention at a high rate is an indication of a more uniform dispersion and better coverage of the active material by particles of carbon.