PLR ( PLASTIC LITHIUM RECHARGEABLE ) BATTERIES USING NANOSCALE MATERIALS : A CONVENIENT ELECTRICAL ENERGY POWER FOR THE FUTURE ?

This communication describes the synthesis of: (i) non toxic and low cost nanocrystalline 
electrode materials which can be advantageously prepared at low temperature; (ii) highly 
conductive electrolyte membranes formed by the nano-encapsulation within a poly 
(acrylonitrile)-based polymer matrix of a solution of LiPF6 in organic solvants. The 
performances of rechargeable PLR (Plastic Lithium Rechargeable) batteries using the 
above mentioned components are presented.


INTRODUCTION
Two events deserve to be reported.
(1) The recent advances in portable electronic devices (portable com- puters, cellular phones, camcorders ) have resulted in a signifi- cant growth in the need for miniaturized power sources able to power these systems.
(2) Mankind is becoming increasingly aware of the damaging health and environmental effects of releasing nauseous gases into the atmosphere due, for instance, to the transportation in big cities.
Consequently high energy and rechargeable power sources are needed for the development of long range electric vehicles leading to the improvement of the air quality [1].
These two demands have motivated the research on electrical energy systems which will be more powerful and occupy a smaller volume than conventional batteries.Obviously these new power sources must be non- polluting and non toxic.Related to that a huge research is in progress in United States, Asia and Europe and three battery systems emerged among the most efficient ones for electric transportation application [1 5].As quoted by Megahed S. and Scrorati B. [1], the first system uses a molten sodium anode, a sulfur-impregnated carbon felt cathode and a solid sodium ion-conducting electrolyte; the second one uses a Li-A1 anode, an iron disulfide cathode, and a molten LiC1-LiBr-KBr electrolyte.However, these two systems operate at high temperature (above 300C) [1]; they are, thereby, inappropriate for the portable electronic devices.That is not the case for the third nonaqueous systems, so called lithium batteries, which can efficiently be used at room temperature and are, therefore, intensively studied at the present time.They are based on a lithium metal or a lithium-inserted graphite as anode, a lithium-ion conducting electrolyte and a cathode material.The cathode material must be both a ionic and electronic conductor and has, generally, a structure which is opened (layered or tunneled) and can, thereby, reversibly intercalate lithium ions.Among the most efficient cathode materials let us mention transition metal dichalcogenides or oxydes such as TiS2, LiNiO2, LiCoO2, LiMn204, V6013 [6-30].
For sake of clarity the working principle of such a Li battery is illustrated on Figure   In Table I are reported the efliciencies at room temperature of Li batteries together with those of the more conventional systems Ni-Cd (nickel cadmium) and Ni-MH (nickel metal hydrure): a lithium battery having a weight of 225 kg is able to power an electric car for 200 km at a speed of 120 km/h; these performances are much lower for the Ni-MH battery and obviously even more lower for the Ni-Cd battery.
Consequently, it appears that the lithium batteries represent (together with the fuel cells [31], which are not the subject of our communication) a leap forward in energy density compared with the conventional systems [32].However, one guesses that a further enhancement of the perfor- mances of the electric vehicles is needed and, therefore, necessitates the production of more powerful batteries which are lighter and occupy a smaller volume than the existing ones.Moreover, regarding the port- able electronic systems quoted above, one guesses also that these systems will become increasingly miniaturized, and this further miniaturiza- tion in portable electronics necessitates a substantial reduction in the weight and size of the lithium batteries.From these two points of view, it will be argued hereafter that the thin and flexible lithium batteries so called PLR (Plastic Lithium Rechargeable) batteries are likely to fulfill the needs of both the electronic market and the electric transportation.In the latter event one can use, indeed, the well known roll-coating technology to get powerful PLR systems which occupy a small volume.On the other hand, in the case of the electronic market one easily guesses that the PLR batteries will be able to fit the shape of the increasingly miniaturized electronic systems and even be an integrated part of it [1].
A PLR battery can be formed by contacting a lithium metal strip, a lithium ion conducting polymeric membrane and a composite cathode membrane obtained by blending the above mentioned active lithium intercalation material with carbon and a polymer binder (Fig. 2).
The Li + conducting polymer electrolyte membrane is, obviously, one of the key components of the battery since the lithium metal is indeed corroded by almost any electrolyte medium.Clearly, the instability of the lithium metal/electrolyte interface is the critical problem which remains to be circumvent in lithium batteries and, therefore, in PLR systems.A way to overcome this problem proposed so far is to replace the metal with carbon-type (coke or graphite) insertion compounds [1, 33].The overall process of such batteries, so called "lithium ion batteries", is illustrated on Figure 3 and can be expressed as [1,33] In fact, the use of LixC6 as lithium source instead of Li metal induces losses in specific energy; moreover, voltage fluctuations, which can expand up to 1.5 V, can occur on cycling [34,35].
In addition, an unnegligeable excess of cathode material is often required in order to get an acceptable cycle life of the "lithium ion batteries", due to the irreversible capacity in the first cycle [1].
For these reasons we have targeted our research on PLR systems which involve a lithium metal film as anode with the objective of reducing: (i) the instability of the lithium metal/electrolyte interface; (ii) the probability of short-circuit which may occur during the charg- ing process due the dendritic growth of lithium metal on the anode.We believed that these two problems can be circumvented owing to an appropriate selection and ratio of the components forming the polymeric electrolyte-membrane.
If the polymeric electrolyte membrane is one of the two key com- ponents of the PLR system, the second one is, obviously, the electrode material.
We will show here also that nanocristalline compounds are promis- ing electrode materials for PLR systems.

CHOICE OF THE MATERIALS-EXPERIMENTAL-DISCUSSION
Choice, Preparation and Properties of the Polymeric Electrolyte Membrane a) Choice of the Materials First of all let us quote that for obvious reasons the polymeric electrolyte membrane must have a room temperature conductivity as close as possible as that of the liquid lithium electrolytes so that the PLR battery will be able to operate efficiently at room temperature.
The main problem with PLR batteries using such polymer elec- trolyte membranes is their inability to operate efficiently at ambient temperature due to the too low room-temperature conductivity (_< 10-6-1cm-1) of the PEO-based electrolytes.Consequently a lot of research has been carried out in order to improve the room tem- perature performances, notably in the ACEP collaboration between Hydro-Quebec in Canada, Elf and CNRS-Grenoble in France [38,39].However, the most successful results so far have been obtained by trapping liquid electrolyte solutions into polymer cages [40].Two distinct methods have been reported: (i) loading the liquid electrolyte into a microporous polymer; (ii) enhancing the viscosity of the liquid electrolyte by adding a soluble (in the electrolyte) polymer until a gel is achieved which can be easily spread onto Teflon moulds leading to polymeric membranes after cooling down at room temperature.The most efficient polymer additives which yield electrolyte-membranes with the required electrical and mechanical properties are poly (methyl methacrylate) (PMMA), poly(vinylidene) fluoride with hexafluoropro- pylene (PVDF-HFP) poly (acrylonitrile) (PAN) [41][42][43][44][45][46][47].We have followed here the above mentioned approach (ii) with the aim of preventing the short-circuit events which may occur during the charging processes of the battery, as quoted above.
We have selected PAN because it is apparently the most stable versus oxidation so that it will not decompose during the charging processes of the PLR batteries [48].Moreover the PAN has polar CN groups which can interact with the liquid electrolyte inhibiting, there- by, its departure.The liquid electrolyte consists of lithium hexafluoro- phosphate (LiPF6) dissolved in an appropriate mixture of ethylene carbonate (EC) and propylene carbonate (PC); the polymer-elec- trolyte interaction would be ensured by the labile hydrogen of PC and the polar CN groups of PAN (Fig. 4).In addition the solvants EC and PC have a high dielectric constant (e'r _ 60) which will ensure an efficient separation of the Li + cations from the PF anions and, therefore, lead to a high room-temperature conductivity.Finally these solvants have a low volatility so that they will allow an easy manu- facturing process of the membrane with an accurate control of the ratio EC/PC.
When we started our research on PLR systems, 5 years ago [48], no one, to our knowledge, had used LiPF6 in onjunction with PAN- PC-EC.The other authors used mainly LiCIO4 and LiAsF6 instead of LiPF6 becaus the C10-and AsF anions are less easily oxidized than PF; therefore the anodic stability of the electrolytes using LiC104 or LiAsF6 would be greater.However we checked that the oxydation potential of PFremained rather high, around 4.3 V vs. Li [48].In fact we have not considered here LiCIO4, due to its explosive character in presence of organic solvants, and LiAsF6, due to the toxicity of Arsenic.Moreover the highest cathodic stability was ob- served for the electrolyte membranes using LiPF6 [48].

b) Manufacturing Process
The manufacturing process of the polymeric electrolyte membrane is illustrated on Figure 5" the reagents were first mixed at the required mass ratio in the inert atmosphere of a dry box and, then, heated at 110C in order to get a solution of appropriate viscosity which can be easily cast onto the substrate.Membranes having a translucent elastomeric appearance were finally obtained after cooling at room temperature.

c) Main Characteristics
One easily guesses that the conductivity's of these hybrid electrolytes can be tailored by a judicious selection of the ratio of the different components.dissolution ofLiPF6 + EC + PC at 25C; drying PAN at 60C (48 hr.) under vacuum /T ---110C (20') @ LiPF6 + EC + PC @ casting>> of the viscous solution teflon mould @ dissolution ofLiPF6 + EC + PC + PAN* A (still odental) addt#ve A s r e tli of t eleewoe membri meal foil inte in t P e.Among the numerous compositions which were studied [48], we established that the highest conductivity was achieved for the composition (molar %): 20 PAN//5 LiPF6/40 EC/35 PC.The conductivity is indeed close to that of liquid electrolytes (Fig. 6), [48].
Moreover the membrane has the required dimensional stability, with a Young modulus of 1.7 MPa at 25C [48].Figure 6 shows that this composition is rather critical: a slight modification of the EC/PC ratio or of the LiPF6 content induces, indeed, a drop of the conduc- tivity, particularly at low temperature.In fact, this drop of the conduc- tivity arises from a crystallization process.The latter is illustrated on Figure 7 which displays the DSC thermograms of various membranes having different proportions of EC and PC.
The peaks A and B which appear only for the membranes 1, 2 and 3 account for a crystallization process (peak A) followed by a melting process of the crystallized species (peak B).It is noteworthy that the membrane 4 which does not undergo any visible crystallization process (Fig. 7) has the highest conductivity reported above (Fig. 6).It corre- sponds, indeed, to the composition (molar %): 20 PAN//5 LiPF6/40 EC/ 35 PC.In fact, the drop of the conductivity observed for the other com- positions (Fig. 6), which exhibit the crystallization process evidenced on Figure 7, would imply that the crystallized species involve Li + ions which are trapped in crystallized EC.On the other hand, for the membrane 4 (Fig. 7), having the highest conductivity, all EC interacts efficiently with PC so that no crystallization of EC occurs.The Figure 8 log.o :.:z .: -130-llO -gO .90-50 -30 -10 10 30 J0 . .

R
Li + e" Li 0 Reversibility ofthe reaction FIGURE 8 Potential sweep, J f(E), of a stainless steel electrode versus Li metal foil sandwiching the membrane* 4 (see Fig. 7) having the highest conductivity (see Fig. 6).*The additive A (see Fig. 5) is added.
illustrates the rather good electrochemical stability of the membrane 4; the electrochemical stability window was determined by running volt- ammetry sweeps of a stainless steel blocking electrode vs. a lithium- foil counter-electrode.
The whole results .wouldimply that the membrane 4 could be advantageously used as electrolyte for efficient PLR systems.
Choice, Preparation and Properties of the Cathode Materials a) Choice of the Materials Commercially available lithium batteries use layered LiCoO2 cathodes [49].However the high cost and toxicity of cobalt motivate the development of cheaper and non-toxic electrode materials.In this prospect nanocrystalline or/and X-ray amorphous, lithium-manga- nese oxides are very attractive alternatives [50][51][52][53]; moreover these materials can be advantageously prepared at low temperature leading to low-cost manufacturing processes.
Ten years ago, we have indeed shown, using many examples, that the control of crystallite size was a key factor which determine the electrochemical performances of the electrodes.The following "electro- chemical model" was deduced [51]: "Nanocrystalline materials are likely to have an enhanced electrochemical activity, compared with that of their microcrystalline homologue, provided that the first electrochemical process which intervenes is a discharge of the Li battery.
The discharge should begin with an electrochemical grafting of the Li + ions at/near the crystallite surface.The structural defects or distorsions at/near the crystallites, which are more obvious in the nano- scale region than in the microscale one, can indeed act as reversible grafting sites for Li + ions.
The following second electrochemical step occurs during the discharge of the Li battery (when the crystallite structure is adapted): intercalation of the Li + ions into the crystallites.
The discharge-charge electrochemical process can be depicted as: grafting intercalation &intercalation degrafting.
These concepts are important because they allow to foresee when it is preferable to use nanocrystalline electrode materials (or micro- crystalline electrode materials), depending on whether the first electrochemical process of the Li battery is a discharge (or a charge)".
We will illustrate, here, these concepts for two different nanocrystal- line electrode materials based on lithium manganese oxide: (i) nanocrystalline spinel-type "LiMn204" [48, 52].(ii) nanocrystalline rock-salt type "LiEMnO3" [50, 48]  An original polymeric route, that we have recently patented [53], has been used.It is depicted on Figure 9.  Homogeneous starting materials have been obtained from an aqueous HNO3 (IN) solution into which was added appropriate amounts of polyacrylamide (PAam, Aldrich M.W. 10.000), lithium carbonate (Aldrich 99,997%) and manganese (II) acetate tetrahydrate (Aldrich 99 + %).First of all a gel is formed and afterwards a xero- gel (hybrid polymer) after removal of the solvent.The controlled explosive oxydo-reduction reaction, occurring at 200C under oxygen atmosphere (Fig. 9)-, of the hybrid polymer leads to ultrafine crystallites (,,50.A,) of the spinel phase of composition Lil_36 ["']35 3,552T.1oct,-, MrI2_28L126] IO4 with S 0.02 [48, 53].Eneggy (eV) FIGURE 10 The average oxidation state of Mn, which is higher than 3.5, and the Mn RO, Mn--Mn coordination numbers, which are lower than 6, observed for the nanocrystalline 'LiMn204' have been deduced from XAS analysis.They were correlated to the higher structural disorder which is characteristic of the nanocrystalline texture [52].For instance, this figure shows the XANES (Mn K edge) spectra for the nano- crystalline LiM n2O4'.
the combination of chemical titration, XPS analysis [48] and XAS analysis [52], (Fig. 10).The TEM micrographs (Fig. 11) and XRD pat- tern (Fig. 12) are characteristic of the nanocrystalline texture with an average crystallite size of 50  (ii) The nanocrystalline rock-salt type electrode prepared at room temperature As quoted by Kim J. and Manthiram A; [50], complex metal oxides are usually prepared by firing of the raw materials at elevated tempera- ture.This approach cannot yield to the accessibility of metastable nanocrystalline or/and X-ray amorphous phases which may be highly attractive as electrode materials for lithium batteries according to Ref.
Here we have used LiI to reduce the permanganate ion (MnVUO-) in Na + containing aqueous medium [50] in order to produce a Mn TM rich nanocrystalline oxide of composition close to LiEMnO3.The detailed preparation process is illustrated on Figure 13.
The as-prepared sample was found to be amorphous to X-ray diffraction as similarly observed by Kim, J. and Manthiram, A. The detailed textural and structural investigation using XPS, XAS, TEM, XRD etc analysis is currently in progress and will be reported later on [55].

ELECTROCHEMICAL PROPERTIES OF THE PLR BATTERIES
The electrochemical measurements were performed using a computer- controlled potentiostat/galvanostat (Tacussel, PGS 201T model) for the PLR cells assembled by sandwiching the polymer electrolyte (membrane 4 in Fig. 7) between the lithium anode foil and the compo- site cathode.
The composite cathode was made according to the preparation procedure proposed by us [48,53] and that proposed by Sun, Y. K. and Jin, S. H. for their efficient PLR cells [56].
The as synthesized nanocrystalline "LiMn204" (or "Li2MnO3") powder was added to PAN solution in demethyl sulfoxide (DMSO) as a solvent.The slurry was spread onto a carbone active tissue of 2000 m2/g surface area (Actitex) and dried under vacuum for 16 h at 110C.The dried composite cathode was then compressed with a roll presser in the inert atmosphere of a dry box.
The PLR cells were assembled, in the dry box, by sandwiching the above mentioned membrane 4 (see Fig. 7) between a lithium foil and the composite cathode.
The reversible capacity corresponds to 0.8 mole of Li per mole of electrode material as shown on the Figures 14 and 15.This rather high and reversible capacity was related to the presence of the structural defects arising from the nanoscale structure, in agreement with the "electrochemical model" mentioned above.The influence of the struc- tural defects has been investigated recently using XAS or Li NMR [52] and also the evolution of the open circuit voltage of the cell as a function of the inserted Li [48].
In conclusion, the PLR batteries based on the electrolyte membrane of composition (molar %): 20 PAN//5 LiPF6/40 EC/35 PC, sand- wiched between a Li foil and nanocrystalline "LiMn204" or "Li2MnO3" composite cathode appears to be highly attractive for industrial point of view, although further investigation is needed.) with the number of cycles for the PLR cells using nanocrystalline 'LiMnzO3'. FIGURE4

FilmFIGURE 5
FIGURE 5 Manufacturing process of the electrolyte membrane.
b) Synthesis Process (i) The nanocrystalline spinel electrode prepared at 200C 10nm

FIGURE 14
FIGURE14 Discharge/charge behavior (C/5) with the number of cycles for the PLR cells using nanocrystalline 'LiMn204'.

FIGURE 15
FIGURE 15 Discharge/charge behavior (C/5) with the number of cycles for the PLR cells using nanocrystalline 'LiMnzO3'. 1 .

TABLE Performance
Specific Energy (Wh/kg) The later was deduced from