For decades most of the research for positive electrode technological achievements was devoted to transition metals oxides such as (M = Co, Ni, Mn). Thus the compounds synthesized from transition metals and poly anions have been actively investigated during the past decade. Polyanions groups such as (PO₄)³⁻ have many advantages such as:

• Structural stability (essential for security and element cycling);

• The chemical nature and their association with redox couple of the transition metal that gives high potential values of those oxides;

• the possibility of varying the crystal structures [2] .

Trad K. et al. [3] demonstrated that the LiFePO₄ phase exhibits interesting electrochemical performances. And various studies have shown that several factors such as the synthesis method, particle morphology, and electrical conductivity of the material are very important for achieving good electrochemical performances in lithium battery. In this work we try to synthesize phosphate based ceramic and increase its conductivity by using a very low cost and accessible product (NPK fertilizers) as the synthesis precursors and as the Al dopant.

In this section, the paper will focus on the different steps of the synthesis of a phosphorus base ceramic with NPK fertilizer, lithium chloride and iron chloride (3) as Li and Fe precursors. The structural and electrical characterization of the obtained materials at room temperature and high temperature will also be discussed.

Figure 2(a) shows the X-ray diffraction pattern for the sample sythesized with the NPK fertilzer. The overall pattern (with the presence of various sharp peaks) indicates a relatively high degree of crystallinity. Diffraction peaks were identified as belonging to two phosphorus based crystalline compounds: iron phosphate, FePO₄ and lithiumiron phosphate, LiFePO₄—“triphylite” [4] . Indeed, the presence of peaks at 21˚, 26˚ and 37˚ indicate the presence of FePO₄ component as well as the presence of peaks at 17˚, 26˚ and 37˚ confirms the presence of LiFePO₄ in the ceramic. However, we note other peaks, showing the possible presence of unidentified compounds which certainly came from the fertilizer chemical elements (N, K, Si, Al, etc). For us the presence of potassium should not have significant influence on the conductivity properties as belonging to the same chemical family as lithium and therefore having the same properties. As against, the presence of nitrogen would be considered as a dopant that could be beneficial.

The results of EDX analysis made for the samples (Figure 3) are consistent with the fact that the ceramic is based on a lithium iron phosphate, with Al, S and Ca atoms as dopants, making it a compound of possible general formula with the presence of Ca and S impurities. The aluminium peak is the sharpest. This could be explained by the presence of an aluminum rate larger than the other compounds. Thus, the ceramic may be rather considered aluminum based materials.

The thermal expansion of the samples was performed with a thermal cycle of 5˚C/min and rising to 20˚C/min without any plateau.

Figure 4 shows the sample behavior under a cycle heat treatment from room temperature to 900˚C. The first phase corresponds to an expansion of the material lattice parameter (Figure 4(a)) which starts from the ambient to 400˚C. This behavior can be explained by the natural behavior (thermal expansion) of any heated material. The second phase is from 400˚C to around 800˚C. This phase corresponds to contractions that are due to different physical phenomena namely dehydration and optionally some chemical compounds extrusion. Our ceramic densification begins only around 820˚C. From 820˚C, the curve shows a significant shrinkage which can be at

tributed to glass transition of the material and thus the presence of a new and more stable crystal structure.

Figure 4(b) shows a more detailed view of our ceramic behavior between room temperature and 800˚C. Four physical phenomena (shrinkages) can be observed. Between 400˚C to 450˚C, heat treatment accelerates the creation of structural defects which is characterized by a contraction of the lattice due to the iron components extrusion. This phenomenon is observed by Julien et al. [5] who detected the presence of Fe₂O₃ and Li₃Fe₂(PO₄)₃. Other shrinkagesat 460˚C, 600˚C and 725˚C could be explained by the presence of other chemical impurities (N, K, Al, etc.) observed on EDX spectra (Figure 3).

Figure 4(c) shows a modeling of the thermal behavior of our ceramic between 30˚C and 400˚C. It is noted that the thermal expansion is a quadratic function of the temperature for this range (before the shrinkage). This relation can be written as:

with, and

The thermal expansion coefficient is determined for the temperature range between 30 and 150˚C at the heating rate and is equal to. This value is 10² times more important compared to values often found in the literature [6] . This low thermal expansion coefficient, compared to LiFePO₄, is an advantage for using this material under extrem conditions.

Figure 5 shows the I-V curves under different temperatures (33˚C, 55˚C, 70˚C and 83˚C). The observed linearity of the curve reflects the ohmic behavior of the material under these temperatures. What could be explained by the fact that in this temperature, the conductivity is dominated by electrons movment. However the resistive or ohmic character of the material disappears when the temperature is increased (55˚C, 70˚C and 83˚C). In fact the vibrational contribution of atoms, at temperature over 33˚C, gives rise to ions which will contribute to the conduction phenomenon. Also when the material is subjected to a temperature increase, the space charges created by the impurities and defects increase the current through the material. This phenomenon, gradually, increases the effect of the ionic conduction which would dominate at high temperatures.

So we can conclude that at low temperature, conductivity is dominated by the electronic conduction while at high temperature the ionic one becomes important. This explains the non-linearity of these curves as soon as the temperature increases.

We modelized the DC current variation as a function of the voltage by a quadratic equation

The constants values are summarized in the Table 1. At low temperature.

These conductivity values obtained using the (Equation (1)) (Table 1) are larger than those usually obtained (in the case of intrinsic LiFePO₄) [7] .

In most semiconductors or insulating materials, the conductivity is generally thermally activated according to the Arrhenius law:

so, (5)

This relation (Equation (5)) permits to calculate the activation energy of this material at this temperature

range. Its value (0.17 eV) is less important than those found in literature (0.31 - 0.5 eV) for LiFePO₄ materials [5] . The activation energies calculated at about 300 K decreases with increasing LiFePO₄ synthesis temperature according to Maccario [8] . The presence of many impurities (K, Si, Li, Fe, etc) can explain this difference and the weakness of the activation energy due to the ionic conduction. The result was expected because the value of ourceramic conductivity is 10² times higher than that of LiFePO₄.

The curve behaviour in the Figure 6 is known under the name of Universal Dynamic Response (UDR) [9] , in which the AC conductivity versus frequencies obeys the following law:

where and are constants at a given temperature. is a measurement of the electric relaxational behavior, usually with a value of: [9] . Many “elementary” hops of the mobile ions are seen (per unit time) within a sufficiently short time window, corresponding to the plateau of the conductivity. Eventually, however, only a small fraction of them prove “successful”, contributing to the DC conductivity and thus constituting the long-time “random” hopping dynamics of the ions. In glassy and structurally disordered crystalline electrolytes, the “elementary” hops as well as the fraction of “successful” ones are usually Arrhenius activated [10] . The particularity of this material is the fact that such behavior is observed in dielectric or insulating matreials at high frequencies while it is observed here at relatively low frequencies (<10⁴ Hz). We used the Equation (4) to determine the values of and with, and found according, at ambient temperature, the following equation:

For Sidebottom [11] , the value of the exponent n is depending to the dimension order in the conducting space.

The value of the conductivity from the UDR curve is appreciably 10 times less important that those found in the DC conductivity measurement.

In this work, we have shown that it is possible to use a very low cost product as (NPK fertilizer) for a LiFePO₄ based ceramics for accumulators. The physicochemical properties by XRD and EDX showed that this material is mainly composed of aluminum, phosphorus, calcium and impurities considered as dopants (Fe, K, Li etc.). A chemical component as is proposed as synthesized ceramic. This material has yielded interesting results for the electrical conductivity (10² times greater than that of LiFePO₄) that is one of the most interesting properties for batteries. On the other hand the low thermal expansion coefficient shows that it may be more advantageous to use this material in certain temperature conditions.

Finally we can conclude that we have achieved an electronic ceramic using very low cost products with interesting electrical and physical properties which could improve the batteries performances.

Advanced studies are in progress with other physico-chemical analysis to determine a quantitative composition of this material.

Financial support of the Senegalese Ministry High Education and Research and the West African Economic and Monetary Union (WAEMU) are gratefully acknowledged.