Lithium battery with a polymer separator

TECHNICAL FIELD [0001] The invention relates to the field of batteries, in particular the lithium-ion type without graphite electrode.

[0002] The invention relates in particular to a battery with a membrane used as a separator between electrodes. This membrane has special physical features that allow it to be used in lithium power batteries, i.e. fast charging and discharging batteries.

BACKGROUND ART [0003] A lithium-ion battery is generally made up of two electrodes separated by an electrolyte allowing the flow of lithium ions from one electrode to another when charging or discharging the battery. The electrodes are made of active material at the electrochemical level and capable of placing lithium in its structure. The most common lithium-ion technology uses graphite on the anode and lithium cobalt oxide on the cathode.

[0004] Regarding electrolytes, two approaches have been developed: some work (see, for example, US-A-4 303 748) has resulted in the use of dry electrolytes which are made of a polymer in the form of a compact membrane in which the lithium salt is incorporated. The polymers used in this case should have in their structure chemical functional groups capable of generating strong interactions with the lithium salt in order to dissociate it and thus enable the transport of lithium ions between the electrodes. As a result of meeting the set safety criteria requiring that each element be in a solid state and in a durable form, the accumulators obtained in this way are, however, not very efficient, especially at ambient temperature: in fact, the ionic conductivity of these systems is very low, up to about 60 ° C . Currently, no action in terms of power at ambient temperature is considered with this type of dry electrolyte: power batteries should be able to charge in a few minutes and to provide strong current peaks when discharging.

[0005] The classic electrolyte structure actually refers to a microporous separator that serves as a liquid electrolyte substrate that locates in the microspores and prevents any contact between the two electrodes that would possibly cause an internal short circuit (Venugopal G. et coll .: «Characterization of microporous separators for lithiumion batteries », Journal of Power Sources 1999; 77: 34-41). For example, liquid electrolytes are made of a mixture of organic solvents of the cyclic and acyclic carbonate type in which the lithium salt, most often lithium hexafluorophosphate (LiPF6), has been dissolved. Modern commercially available separators are microporous polyethylene (PE) and / or polypropylene (PP) membranes, the degree of which₂ porosity ranges from 30 to 50%, with a thickness of about 25 μm and a pore size of 100x40 nm. Polyvinylidene fluoride (PVdF), optionally copolymerized with propylene hexafluoride (HFP) has also been used.

[0006] On the other hand, modern commercially available batteries charge a minimum of an hour. Well, for some applications called high power applications it is desirable that the batteries quickly charge and discharge.

[0007] The use of graphite on a negative electrode inhibits any high power application: in fact, during the passage of strong charging currents, lithium metal dendrites form on the anode surface, which causes an internal short in the event of contact between two electrodes via these dendrites. This is due to the low placement potential (100 mV / Li⁺/ Li) Li ions⁺ in graphite. To remedy this inconvenience, low porosity membranes or compact and plasticized membranes have been developed, as described above.

[0008] At the same time, it was suggested to replace graphite with lithium titanium oxide Li4Ti5O12: Li placement potential⁺ in this material (1.55 V / Li⁺/ Li) is slightly higher than the graphite potential, which eliminates any risk of dendrite formation, even under the influence of strong currents (Nakahara K. et coll .: «Preparation of particulate Li4Ti5O12 having excellent characteristics as an electrode active material for power storage cells », Journal cf Power Sources 2003; 117: 131-136). However, membranes have not been optimized for these new electrodes. In particular, the lack of "binding" interaction and compatibility between these electrodes and the microporous separator results in mechanical malfunction of the assembly.

PRESENTATION OF THE INVENTION [0009] The main object of the invention is to provide a lithium-ion battery that would not use a graphite anode.

[0010] The invention relates to a battery comprising a positive electrode, a negative electrode from a material with a lithium placement potential above 500mV / Li⁺/ Li, as a separator, a membrane with a tortuosity less than 2 containing a fluorinated polymer, whose thickness is between 100 and 120 μm and a porosity between 50 and 90%, and liquid electrolyte or in the form of a gel in the separator.

[0011] Preferably, the fluorinated membrane used as the separator contains pores with a size comprised between 1 and 10 μm, i.e. very large compared to current applications.

[0012] Preferably, the membrane is made of polyvinylidene fluoride, preferably co-polymerized with propylene hexafluoride.

[0013] The pores preferably have the shape of transverse longitudinal channels.

[0014] The membrane is used in conjunction with a liquid or gelled electrolyte containing, for example, LiPF6, in a battery with an electrode whose Li placement potential is higher or equal to mV / Li⁺/ Li, for example in lithium titanate, Li4Ti5O12.

BRIEF DESCRIPTION OF THE DRAWING [0015] Fig. 1 shows the structure of a membrane according to one embodiment of the invention.

DETAILED DESCRIPTION OF PARTICULAR METHODS OF IMPLEMENTATION [0016] Due to the results described in the prior art regarding "dry" electrolytes, the membrane is intended to be used as a separator serving as a substrate for liquid or gelled electrolyte. In order to overcome the above-mentioned problems, it is further desirable to increase the mechanical cohesion between the electrodes and the membrane. Such cohesion is achieved due to the affinity of the membrane with the binder, classically a fluorinated polymer used to make electrodes, here especially with Li4Ti5O12. A fluorinated polymer, homopolymer or preferably a copolymer has therefore been chosen to make the porous membrane of the invention.

[0017] Such a membrane has "hybrid" characteristics, ie between microporous separators and dry polymer membranes. The electrolyte, preferably based on LiPF6, is partly located in the microporosity and partly also in the mass of the membrane because it is partly soluble. So there are three phases:

dry polymer phase;

- gel phase containing a polymer, organic solvent and salt;

- liquid phase in the pores.

[0018] The use of this membrane as a separator therefore allows for a three-phase liquid / gel / solid system due to the addition of a liquid electrolyte, which system has good ionic conductivity, good mechanical strength, excellent cohesion between the membrane and electrodes and a very low risk of leakage.

[0019] The electrical resistance of such a hybrid system is expressed solely depending on

T of the liquid electrolyte resistance according to the formula: R _t = R_r x— in which R_syst is syst liq resistance _e hybrid membrane system, Rliq means the resistance of the liquid electrolyte itself used to form the gel, ε means the porosity of the membrane and T means the tortuosity of the membrane (tortuosity characterizes the actual path traversed by the ion in relation to the distance between the two electrodes).

[0020] It has been found that for proper operation with high power, i.e. with fast charging and discharging, the resistance of the electrolyte system should be as low as possible. Thus, it is preferable that the porosity ε tends to 1, with this 100% electrolyte being only in the liquid state and therefore no longer able to physically separate the two electrodes.

In the same way, the tortuosity T is as small as possible, as close to 1 as possible, although also in this case, the membrane is still porous and ensures physical separation of the electrodes.

[0021] In order to meet these conditions, the polymer membrane according to one preferred embodiment has a high porosity of 50 to 90%, preferably above 60%, while providing good mechanical strength and has pores in the shape of straight transverse channels and less than or equal tortuosity 2.

[0022] Furthermore, the resistance of the entire electrolyte system is related to self-conductivity according to the equation: R_syst

1e

-x—, in which σ ^ is the conductivity of the system ^s syst ^S electrolytic system, e means the distance between two electrodes and S means the surface of the working electrode.

[0023] In order to reduce the resistance, therefore, the surface of the electrodes increases while taking into account the limitations associated with the battery design and the electrodes approach each other in a classic way to reduce the e value, i.e. the thickness of the membrane decreases. In fact, the first membranes made for the invention were thin, typically between 20 and 50 μm, as indicated in the prior art.

[0024] It turned out that better potential parameters were obtained with thick membranes from 60-70 to 120 μm.

[0025] This unexpected result was explained by means of a diffusion mechanism that took place in the electrolyte. The flow of lithium ions between two electrodes is the cumulative effect of migration and diffusion (there is no convection): migration is caused by an electric potential gradient and diffusion is caused by a concentration gradient. The lithium molar stream density NLi is given by the formula: N_Li = N_L^mand + N_L^dand.

[0026] For strong currents when using batteries with high power, Li ions⁺ should be quickly accessible on the surface of the reducing electrode. It turns out that the use of a thick membrane allows, within the scope of the invention, to create a sufficient ion reserve to ensure that lithium is placed on the electrode for strong current peaks. Ion diffusion is therefore high enough that it can then ensure that this supply is discharged during battery operation.

[0027] The solution proposed for the membrane for the battery separator, in particular with the lithium titanate anode, is therefore based on the use of a highly microporous fluorinated membrane (50 to 90%, preferably about 60-70%) and thick (on the order of 100 μm). Such a membrane can be obtained from a fluorinated polyvinylidene fluoride (PVdF) type polymer including a large family of PVdF polymers polymerized with propylene hexafluoride (PVdF / HFP), using a classical method of making membranes.

[0028] Furthermore, the pore size of the membrane is preferably above micron. This increase causes a decrease in tortuosity and thus promotes migration and diffusion of Li ions⁺ in the electrolyte.

Implementation example [0029] The membrane was made using phase inversion technology, which is based on four stages:

(i) The polymer, PVdF / HFP is dissolved in one of its solvents to obtain a solution with an apparent viscosity of about 1800 cP, for example from N-methylpyrrolidone (NMP).

(ii) Then, this solution is coated on a rigid, flat and not very sticky substrate with a plaice, i.e. a film applicator with adjustable gap height and in which the gap thickness is controlled, ideally up to 500 μm.

(iii) The coated substrate is hardened for 5 to 10 minutes, at ambient temperature, in a bath of a solution miscible with the first solvent but in which the polymer does not dissolve; this solution may for example be of the alcohol type, in particular ethanol.

(iv) After polymer coagulation, the membrane plate is placed in an oven for drying for 24 hours at about 50 to 60 ° C.

[0030] The membrane thus obtained represents a preferred embodiment of this invention, with a porosity of between 50 and 90%, a thickness of between 60 and 120 μm, a curvature of less than 2 and in which the pore size can be from 1 to 10 μm. The membrane has no skin on the surface and the porosity is open, as shown in figure 1.

Comparative Example [0031] The membrane thus obtained is compared to a commercially available membrane based _® Celgard polyolefins^®. Two membranes were tested in a battery with a negative electrode made of lithium titanate Li4Ti5O12, with a unit capacity of 160 mAh / g and a potential of 1.55 V / Li⁺/ Li and with a positive electrode made of lithium transition metal oxide, here with LiMn2O4 manganese oxide with a capacity of 120 mAh / g and a potential of 4.1 V / Li⁺/ Li;

Titanate is a deficiency in comparison with manganese oxide. The negative electrode weight is₂

0.27 mAh / cm² and the battery capacity is 0.4 mAh. Liquid electrolyte used in_® combined with membranes is LP30 supplied by Merck^®, composition EC / DMC 1/1, LiPF61M.

[0032] The results are shown in the following table:

Table I: Features and parameters of a battery with a commercially available membrane and with the membrane according to the invention

<td></td><td>Celgard® 2400</td><td>Membrane according to the invention</td>

<td>Polymer</td><td>PP</td><td>PVdF / HFP</td>

<td>Thickness ^ m)</td><td> 25</td><td> 100</td>

<td>Porosity (%)</td><td> 37</td><td> 65</td>

<td>Tortuosity</td><td> 2,4</td><td> 1,9</td>

<td>Pore ​​size ^ m)</td><td> 0,1</td><td> 5</td>

<td>Recovered capacity at 20C (% relative to C)</td><td> 65</td><td> 85</td>

[0033] For a charging system at 20C (full battery charging with DC for 3 minutes it is _®

1/20 h), the loss of capacity is greater with a Celgard membrane^® than with a made membrane. Thus, a battery using a fluorinated microporous membrane as a separator can be recharged to 85% and above its initial capacity in 3 minutes, with a maximum of 65% for a commercially available membrane; the parameters obtained for high power are better than in commercially available microporous membranes.