This project examines whether the basic undoped spinel-oxide material LiNi_.5Mn_1.5O₄ and new, extended, doped variations of LiNi_0.4X_0.1Mn_1.5O₄ (X= Cr, Zn, Fe, Al, Mg, Ga, V, Cu) may be synthesized and operated as a new 5 V spinel battery. A doped material is one that has had impurities deliberately added in order to change its electrochemical properties. In this case, the X cations Cr, Zn, Fe, Al, Mg, Ga, V, and Cu were chosen to test a variety of oxidation states and coordination geometries.

LiNi_0.4X_0.1Mn_1.5O₄ spinel cathodes and unsubstituted LiNi_.5Mn_1.5O₄ cathodes were produced, tested, and compared. This paper presents a revised method for producing experimental cathodes of LiNi_0.4X_0.1Mn_1.5O₄ (X= Cr, Zn, Fe, Al, Mg, Ga, V, Cu) without the impurity phase NiO observed in previous work. It further presents new data on the electrochemical characteristics of these materials. The major purpose of this project is to formulate a new pure 5 V lithium spinel cathode that will produce a more stable battery with a higher gravimetric capacity, and therefore larger energy storage.

The solution was heated to boiling and then cooled to room temperature. It was heated again and allowed to evaporate until about 50 ml of the solution remained. The solution was then put onto a rotary evaporator. The Roto-Vap rotated the solution within a flask, immersed in an elevated-temperature water bath while a vacuum was applied. Under vacuum, the solution boiled at lower temperatures, allowing it to maintain a liquid/gel state. The resulting gel was then put into a furnace at 450° C for 12 hours in air, not including the time it took the oven to reach the desired temperature or the time needed for it to cool down. The gel then became a powdery solid, which was first ground, then kept at 900° C for 16 hours.

A method was then used to create another batch of experimental cathodes, using manganese acetate, nickel nitrate, lithium nitrate, and ammonium hydroxide as the starting ingredients. This method was identical to the first, except that the solution was boiled only once, instead of twice, and allowed to cool.

For physical characterization analysis, a sample was taken from each of the three batches produced, and an X-ray diffraction (XRD) was administered. As the second method produced the best XRD results, only the powders produced by this method were used to construct the experimental cathodes described below.

After the spinels were successfully produced, they were mixed with graphite SFG-6, acetylene carbon black, and a binder (PVDF) to form a viscous cathode solution, which was 84% active spinel, 4% graphite, 4% acetylene black, and 8% binder (by weight). For most of the laminates, about 2 g of the active spinel were mixed. The compound was flattened and smoothed onto a piece of aluminum using a 200-micron blade. The as-casted laminates were placed on glass and put into a low-heat oven (70° C) for 4 hours to overnight in air in order to dry the laminate.

The laminate was subsequently pressed so that it was even at every corner. Then, multiple round cathode discs were produced using a 9/16-inch hole-punch. The discs, along with a blank disc and a sheet of separator, were put into a vacuum oven for 6 hours to overnight. The separator for the battery cell was punched out using a 5/8-inch punch.

A piece of lithium metal was used as the anode, and the produced spinel disc was used as the cathode. After the battery cell was constructed, it was separated to minimize the possibility shorting out. The battery cells, which look like watch batteries, are 20 mm in width, 3.2 mm in height, and are known as 2032 cells.

These cells were then cycled using an automatic battery tester (MACCOR) and a galvanostatic program that tests mainly their various capacities and voltages in charge and discharge cycles.

Figure 1. X-ray diffraction (XRD) pattern of the LiNi0.5Mn1.5O4 spinel obtained after various attempts to eliminate the presence of the NiO impurity. This spinel was made using a lithium nitrate compound. The sharp spikes in the diagram indicate that the sample was a pure single-phase spinel. (Click to view enlarged image)

The voltage profile for a complete cycle of the standard undoped LiNi_0.5Mn_1.5O₄ spinel cell shows two plateaus of voltage through time: one at about 3.97 V and one at about 4.65 V to 4.7 V (Figure 2). The voltage profile for one complete cycle of the doped cell, LiNi_0.4Cr_0.1Mn_1.5O₄, shows three plateaus: one at about 4.01 V, one at 4.65 V, and another at about 4.74 V (Figure 3). The two highest plateaus of the doped cell voltage profile are parallel and close to each other.

Figure 2. Voltage profile for one complete cycle of the undoped spinel cell, LiNi0.5Mn1.5O4. The graph shows the voltage per time for the 10th cycle, illustrating the two plateaus.(Click to view enlarged image)	Figure 3. Voltage profile graph of one complete cycle of the LiNi0.4Cr0.1Mn1.5O4 spinel Li coin cell.(Click to view enlarged image)

Specific capacities of the undoped and doped cells differ. The undoped cell, LiNi_0.5Mn_1.5O₄, shows a slightly lower specific capacity (Figure 4) than the doped cell, LiNi_0.4Cr_0.1Mn_1.5O₄, (Figure 5); however, the undoped cell does show a much more consistent cycling capacity (Figure 4).

Figure 4. Voltage vs. capacity graph of the LiNi0.5Mn1.5O4 spinel coin cell. This plot shows the charging and discharging voltage of the cell through 15 cycles, and indicates a very strong consistency.(Click to view enlarged image)	Figure 5. Voltage vs. capacity graph for the LiNi0.4Cr0.1Mn1.5O4 spinel coin cell. The first charge and discharge of the test seem to show high irreversible capacity. Cycles 2 through 8 graph very close to one another, showing great cycling stability.(Click to view enlarged image)

Measured coulombic efficiencies for the undoped and doped cells also differ. For the LiNi_0.5Mn_1.5O₄ cell, the coulombic efficiency on the ninth cycle is 94.8%, and the discharge capacity is 128 mAh/g, which is 87.1% of the theoretical value (Figure 6). The coulombic efficiency of the LiNi_0.4Cr_0.1Mn_1.5O₄ coin on the ninth cycle was measured as 97.8%, and the discharge capacity is 132 mAh/g, which is 89.7% of theoretical (Figure 7).

Figure 6. Voltage vs. capacity graph of the LiNi0.5Mn1.5O4 spinel coin cell. This plot shows the charging and discharging voltage of the cell through 15 cycles, and indicates a very strong consistency.(Click to view enlarged image)	Figure 7. Capacity graph of the LiNi0.4Cr0.1Mn1.5O4 spinel Li coin cell. Both the charge and discharge capacities merge as the cycles increase. The coulombic efficiency on the ninth cycle is 97.8%, and the discharge capacity is 132 mAh/g, which is 89.7% of theoretical. This material shows about a 3% improvement in capacity over the non-doped standard LiNi0.5Mn1.5O4 spinel (Figure 4), at least after nine cycles. .(Click to view enlarged image)

Chromium, on the other hand, produced near-perfect electrochemical results. Through each charge and discharge cycle of the chromium spinel cell, the voltage and capacity remained correlated (Figure 5), illustrating the cell=s stability. The cell maintained constant capacity through multiple cycles (up to about 30), as illustrated in the identical high capacity cycles in Figure 7. This improved cyclability may be due to the ability of Cr ions to go between Cr(III) and Cr(VI) sites (Balasubramanian et al. 2002).

The voltage profiles for both the undoped and doped cells show interesting plateaus (Figures 2 and 3). In the undoped cell, the first plateau shows the two-electron reduction of Ni⁴⁺ to Ni²⁺ in successive steps. The second probably indicates the presence of some residual Mn⁴⁺ to Mn³⁺ reduction. The doped cell shows three plateaus. One is at about 4.74 V and another at about 4.65 V during discharge. These two show the two-electron reduction of Ni⁴⁺ to Ni²⁺ in successive steps. The third plateau, at about 4.01 V, is smaller and indicates the presence of either residual Mn⁴⁺ to Mn³⁺ reduction, or the doped Cr⁶⁺ to Cr³⁺ reduction (Hong and Sun 2002). The average voltage of the cell is higher than that of the non-doped standard LiNi_0.5Mn_1.5O₄ spinel, which indicates that the stored energy of LiNi_0.4Cr_0.1Mn_1.5O₄ cathode is higher than the cathode of LiNi_0.5Mn_1.5O₄.

The voltage vs. capacity graph of the LiNi_0.5Mn_1.5O₄ spinel coin cell indicates a very strong consistency (Figure 4); however, the LiNi_0.4Cr_0.1Mn_1.5O₄ spinel coin cell shows greater cycling stability in addition to a strong consistency (Figure 5). Also, the first charge and discharge of the test of the LiNi_0.4Cr_0.1Mn_1.5O₄ spinel coin cell seem to show high irreversible capacity. This material shows about a 3% improvement in capacity over the non-doped standard LiNi_0.5Mn_1.5O₄ spinel (Figures 6 and 7), at least after nine cycles. This indicates that the LiNi_0.4Cr_0.1Mn_1.5O₄ cell maintains a more constant capacity than the undoped cell, and does not significantly lose capacity as the cell charges and discharges through multiple cycles.

The LiNi₀₅Mn_1.5O₄ spinel is an excellent cathode material for 5 V batteries; however, synthesis and electrochemical characterization of LiNi_0.4Cr_0.1Mn_1.5O₄ has shown it to be even better. It shows a greater stability with less fade during both charging and discharging of the coin cell. Though all the X metal elements induced into the spinel have not yet been made into cells, the ones that have been developed have shown much promise (i.e., the excellent electrochemical results from the chromium spinel cells).

The next step in the process is to test the remaining doped spinels. After that, research should be conducted to dope the standard spinel with multiple elements to maximize capacity and efficiency at both room temperature and at 50° C. If these batteries consistently sustain high capacity and maintain irreversible capacity at a minimum 50° C, they could become the next energy source for everything from toys to computers to cars. Their high voltage, economical chemical foundations, and affordable materials make this more than possible. The 5 V lithium battery may mark a turning point in the development of a new, revolutionary energy source.

I would finally like to thank everyone at the CMT for being so kind and making me feel right at home.

The research described in this paper was carried out at the Argonne National laboratory, a national scientific user facility sponsored by the United States Department of Energy.