In the current world of technology, efficient connectivity greatly depends on mobility. Portable electronic devices which have evolved almost as our electronic organs thrive on rechargeable power sources―currently on lithium ion batteries (LIBs). The construction of such batteries utilizes Lithium compounds―Lithium Cobalt Oxide (LiCoO2) being the most popular of them all―as active cathode materials. LiCoO2 acts as storage of electricity while charging and as a source while discharging. This costly component turns into the major environmental danger after the service life of the battery if improperly disposed  . The cobalt in such materials posses threats to the ecology being a heavy metal. Lithium being explosive in nature in the elemental form carries the risk of accidental explosion and toxic gas emission while burnt off or ill-treated during informal recycling   .
LIBs find their use commonly in laptop computers, cellular phones, camera, rechargeable lights and numerous modern life appliances and very recently in electric vehicles. An estimated total production of 12.7 billion mobile phones, 94.4 million laptop computers, and 768.9 million digital cameras  was reported by the United Nations till 2010. Clearly equal number of LIBs is to be handled after their lifetime. For the year 2016 alone, 4.7 billion unique mobile phone subscribers were reported  . Estimated 135.98 million  domestic mobile phone subscribers were present in mid 2017, which forecasts the number of LIBs joining the waste stream very soon. Approximately 0.3 million  computers are being consumed each year in the country, of which a substantial amount is the laptop computers having LIBs.
The importance of extractable metals from LIBs for a country like Bangladesh with no primary metallic sources cannot possibly be over exaggerated, while sorted electronic waste stream has already found its way to efficient recycling facilities abroad. In addition, if the key ingredient―LiCoO2―can be recycled locally, domestic manufacturing of LIB could be possible empowering national economy.
From the late ninety’s, researchers have been working on the development of recycling opportunities from LIBs. Thermal  and Mechanical  treatments have been experimented successfully. Solvent extraction by PC-88A  , Acorga M5640  and Cyanex 272   have yielded great recovery rates, however, the availability and post treatment for such solvents along with the cost associated needs much consideration. Although dissolution by N-methylpyrrolidone (NMP) has been proven   helpful, the authors found difficulties with the output (trials with DMSO were also unsuccessful). Chemical leaching with organic   or inorganic   acids and subsequent precipitation of metallic salts to be used as precursors for the formation of active cathode materials- deems to be the most attractive option till today due to the low cost associated and process simplicity. The current work utilizes HCl leaching of LiCoO2 aided by H2O2 as reducing agent after laborious yet simple manual dismantling of lithium ion batteries from laptop computers. Cobalt and lithium were then precipitated prudentially by basic treatments as Co(OH)2 (later turned to Co3O4) and Li2(CO3). Finally they were mixed and calcined to produce LiCoO2.
2. Materials and Methods
Waste laptop batteries (Brand: HP, Type: Li ion, 6 cell) were collected from the local scrap market (Elephant road, Dhanmondi, Dhaka). The chemical reagents (HCl, NaOH, H2O2 and Na2CO3) used in this study were of analytical grades (Merck, Germany and Scherlue, Spain). For all purposes de-ionized water (pH 6.5 - 7.5) was used.
For the estimation of carbon and organic contents, a CHNS-O (Brand: Thermo Fisher, Model: Flash 2000, Origin: USA) was used. To identify the metallic elements in the active electrode materials, a WDXRF (Brand: Shimadzu, Model: LAB Center XRF-1800, Origin: Japan) was used. Phase identification data of several intermediate products and raw materials were obtained using XRD (Brand: Bruker, Model: D8 Advance, Origin: Germany). Concentration of ions in different intermediate solutions was characterized by an AAS (Brand: Shimadzu, Model: AA7000, Origin: Japan). A (Brand: Jeol, Model: 71031SE2A, Origin: Japan) FESEM was used to acquire micrographs.
3.1. Mechanical Breaking
Waste laptop batteries were mechanically broken with pliers to separate the plastic casing, connectors, thermo-couples and additional materials from the 6 cell assembly. Each cell was discharged by dipping in a 5% NaCl solution for 2 hours. The cells were then broke mechanically under a fume hood with pliers and a screw driver to unravel the coiled PE separators and the electrodes to be kept on a steel tray inside the fume hood to naturally dry for 3 days.
3.2. Separation of Aluminum and Active Electrode Materials
The Cu electrodes and PE separators were then manually scraped to remove active electrode materials (AEM, consisting graphite and LiCoO2) as far as possible. Further scraping was done by the wet brushing of PE separators with DM water and a brush.
The Al electrodes and attached AEM were treated with 1M 500 ml NaOH. The Al foil reacted with the NaOH to form Na2Al2O4 and the AEM got separated. This portion of AEM was treated again with 1 M 100 ml NaOH to assure any un-reacted Al got in the solution. The solution (total 600 ml) was then filtered to recover the AEM and washed subsequently with DM water before drying at 100˚C overnight. The filtrate was treated with dry CO2 bubbling through it with rigorous stirring. Eventually the solution turned white and Al(OH)3 precipitation formed. It was observed that this process was accelerated in warm (~50˚C) condition. It was then filtered and washed before drying overnight at 100˚C. The filtrate is again treated with dry CO2 to recover additional Al(OH)3 precipitates. The amount achieved from this second treatment was reasonably small. It was then filtered and dried the same way. A portion from the filtrate was taken for Atomic Absorption Spectrometer (AAS) to quantify any remaining amount of Al. The dried Al(OH)3 was characterized with X-Ray Diffractometer to identify any possible impurities and confirm the Al(OH)3 phase(s) formed. It was finally treated at 600˚C for 4 hours to get Al2O3. The separated AEM were ball milled to fines. The milled AEM was analyzed in an elemental analyzer (CHNS-O) to quantify the amount of graphite and organic materials remaining. Also X-Ray fluorescence was done to reconfirm the type of active cathode material.
The leaching and simultaneous removal of graphite from the AEM was done by 3 M HCl with a solid to liquid ration of 1:20, aided by 5% H2O2. The solution was then filtered to acquire the graphite particles on the filter paper to be washed and dried subsequently. The dried graphite powder was then analyzed by XRD to confirm the absence of LiCoO2.
3.4. Cobalt and Lithium Recovery
A portion of this solution was analyzed in the AAS to estimate the amount of leached Co and Li. Rest of the solution was treated with 65 ml 4 M NaOH (per 100 ml leach solution) to reclaim Co as cobalt hydroxide precipitates at near pH 11. The Co(OH)2 obtained was then treated at 600˚C for 3 hour to get Co3O4. The remaining solution was condensed to half of the initial volume by heating. Later it was treated with saturated Na2(CO3) and boiled for some time to get a precipitate of Li2CO3. The solutions-before and after the Li separation-were also taken for AAS analysis to measure the Co and Li recoveries respectively.
3.5. LiCoO2 Formation
To synthesize LiCoO2 stoichiometric proportions of Co3O4 and Li2CO3 was calculated to be 1:1.5 (molar) according to the following equation:
A set of samples of total mass 1 gm was prepared by mixing the weighted reactants in a mortar-pestle and pressing in a die up to 20 kpa to form green tablets. The tablets were then sintered in a tube furnace with and without air supply. The sintering cycles were similar to Figure 1.
The reaction suggests that, O2 is required for the fulfillment of the reaction. Hence, later experiments were done with an additional air flow provided by a pump attached with the tube furnace. The tablets were characterized by XRD after each time period of sintering, to identify the phases formed.
To evaluate the recovery process and possible electrode material loss, the physical dimensions of the electrode current collectors and the PP separators were measured (Figure 2).
Figure 1. Sintering cycle for the synthesis of LiCoO2.
Figure 2. Separators (top) and Electrodes (bottom) of LIB.
Table 1. Measurements of the electrodes and separator.
Table 2. Amount of different constituents of a LIB.
4.1. Active Electrode Material Identification
The active electrode materials were identified by X-Ray Fluorescence (XRF) analysis (Table 3). It gave a definite indication that the active cathode material
Table 3. Amount of different elements in the AEM.
was rich in Co, later identified to be LiCoO2. Lithium cannot be effectively identified by XRF, being a lighter (atomic no: 3) metal. There was an indication of the presence of small amounts of Cu and Al in the cathode material, possibly due to corrosion and the dismantling-scraping process.
Also the content of graphite was estimated with an Elemental (C, H, N, S, O) Analyzer. The presence of insignificant amounts of H, N, S and O indicated that almost all organic materials used were volatile and the remaining was mainly carbon (graphite). From this result, the amount of LiCoO2 in the electrode material was estimated to be 54% (round figure) to be used in the leaching calculations.
4.2. Reclamation after Leaching
The Atomic Absorption Spectroscopy results are Table 4.
It depicts the concentration of metallic ions in solutions achieved in different steps. The phases of the reclamation products were identified by XRD later on.
4.3. XRD Analysis
The XRD analysis of active electrode materials (Figure 3) detected the presence of Graphite (JCPDS: 65-6212) and LiCoO2. After the removal of Al current collector by NaOH, subsequent CO2 treatment and drying; Al(OH)3 (as Bayerite, JCPDS: 20-0011) was formed, which turned to Al2O3 (JCPDS: 50-0741) after calcination at 1000˚C for 1 hour (Figure 4). The residue after leaching was almost pure Graphite (JCPDS: 65-6212). The reclamation products were identified to be Co(OH)2 (later calcined to Co3O4 (JCPDS: 73-1701) at 650˚C for 1 hour) and Li2CO3 (as Zabuyelite, JCPDS: 83-1454) as shown in Figure 5 and Figure 6. Co(OH)2 was amorphous in nature, as a result no crystalline pattern was achieved in the XRD.
The formation of LiCoO2 at 950˚C for different time durations (6, 12, 18 hours) showed different scenarios (Figure 7). Without air the 6 hour treatment showed the formation of less crystalline LiCoO2 (JCPDS: 50-0653) and an impurity phase, possibly LiAlO2 (JCPDS: 33-0776)-formed in contact with the porcelain boat. After the improvisation of additional Al2O3 layer below the green tablet, the 12 hour treatment showed much crystalline LiCoO2 and Co3O4 (JCPDS: 73-1701) and some unidentifiable impurity phase-as an indication of Li2CO3 loss and incomplete reaction. The 18 hour treatment resulted in LiCoO2, Co3O4 and CoO (JCPDS: 77-7548). At this point it seemed conclusive that further treatment will never reach a complete conversion to LiCoO2, as there is a Li deficiency and a possible tendency of Co3O4 to convert to CoO with time under the treatment conditions.
Figure 3. XRD spectra of LIB active cathode materials before and after leaching.
Figure 4. XRD spectra of aluminum extraction as alumina.
Figure 5. XRD spectra of cobalt oxide synthesis.
Figure 6. XRD spectra of lithium carbonate synthesis.
Figure 7. XRD spectra of LiCoO2 synthesis without air.
Table 4. Normalized AAS data of solutions derived at different stages.
With air, the 6 hour treatment at 950˚C showed complete crystalline formation of LiCoO2 (JCPDS: 50-0653) with no observable impurities (Figure 8). The product remained the same in the later experiments with prolonged time durations to 12 and 18 hours.
4.4. SEM Micrographs
The SEM micrographs shown in Figure 9 depicted irregular shaped grains of synthesized LiCoO2 particles throughout the sintered body.
The dismantling process proved to be an effective one, regardless the labor intensive procedure requiring very little sophistication. Needless to mention, the safety requirements for the dismantling process-fume hood, safety goggles, gloves, etc. were maintained as required.
It was observed that the discharging step of the cells in brine solution initiated some leakage (through the cell vent) to the brine solution and turned the solution color to brown/orange. After some time, the solute precipitates itself and could be effectively filtered. It was assumed to be the commonly used electrolyte (LiPF6) dissolved in Ethylene/Dimethyl Carbonate (EC/DMC) solvent, but the assumption could not be verified as the amount of the filtrate was very small and it projected no crystalline peaks in XRD.
It was found that aqueous stirring or ultrasonication had poor effect on the separation of electrode materials from the metallic electrodes and the polymeric separators. Even when DMSO solvent used for PVDF binder removal, ultrasonication resulted in a separation of small metallic particles. As a result scraping was deemed to be the optimistic way.
Figure 8. XRD spectra of LiCoO2 Synthesis with air.
Figure 9. SEM micrographs of synthesized LiCoO2.
The separation was very easy for the Cu electrodes because it contained mainly graphite on both sides, unlike the Al electrodes containing graphite as well as LiCoO2 pasted with organic binder(s). As for the PE separators, the scraping process was intermediately successful leaving a good amount of graphite and LiCoO2 still attached on both side. It was later wet scraped. Aluminum Electrodes―in most of the cases―got highly corroded and blistered; possibly due to rapid charge/discharge of the cell assembly in actual use. Also the PE separators got black shades in some cases, on the LiCoO2 side due to the same reason. At some points Al electrodes and AEM got severely attached with the PE separators, possibly because of the severe heat generated due to charging/discharging rapidity. Hence brushing did not help the separation of AEM from Al electrodes. And negligible amounts of AEM were stuck with the PE separators even after brushing treatment.
The mechanical breakage of the cell assembly and later the cells, readily provided reusable or recyclable Cu foils, PE separators, steel casings and plastic (PE, PP, Bakelite, etc.) portions. After the leaching process, the almost pure graphite was achieved which also could find numerous applications. The NaOH treatment of the cathode dissolved the Al and was later recovered by the CO2 bubbling. The calcined Al(OH)3 turned to Al2O3 which has more commercial value.
During experiments it was found that without milling complete leaching of LiCoO2 is tough as it gets trapped inside the graphite. The sheet like layered structure of graphite might also contribute to this effect by trapping Li+ ions inside from both the electrolyte and the reaction product of the AEM from charging/discharging cycle. Hence the AEM was milled to fines.
The AAS data confirms that the aluminum current collector has been effectively removed from the cathode. For Cobalt and Lithium ions derived from the leaching of LiCoO2, 99% Co and 30% Li could be successfully recovered by the leaching and reclamation process. The Li recovery was low due to some process factors. The solubility of Li2CO3 in water decreases with increasing temperature. There was no clear indication of completion of the formation of Li2CO3 at near 100˚C. Subsequent filtering also incorporated the solidification of some NaCl crystals on the filter paper alongside Li2CO3, as the solution was already very rich in Na+ and Cl− ions. The precipitates were washed at least 5 times with boiling DI water to remove the NaCl. Collaterally some Li2CO3 were lost in the process. Although the recovery of Li2CO3 was less, the purity (confirmed by XRD) was very high.
The formation of LiCoO2 depends on various parameters like composition, environment and heating rate, etc.. Researchers   have formed LiCoO2 from various raw materials at above 800˚C with different time duration (close to 24 hrs). A simple sintering cycle was hence chosen to verify the outcomes.
The addition of air supply with the tube furnace provided tremendous results. The formation of LiCoO2 was complete in all the samples. It could be suggested that under such conditions LiCoO2 formation may be complete even in less time. The authors hope to perform electrochemical tests in order to measure the reversible capacity of the re-synthesized LiCoO2 as future research. It was observed that porcelain boats (made of Kaolinite clay, a form of alumino-silicates) reacts with Lithium salts to produce aluminates or silicates if there is a contact with the green tablet. Hence, a layer of α-Al2O3powder was put on the porcelain boats bellow the green tablet(s) later on.
In this study, a lithium ion battery assembly has been manually broke into constituent electrodes, separators and other components. A proven reclamation process of Li and Co has been investigated by leaching with HCl and precipitating Li and Co compounds by basic treatments. The compounds (Li2(CO3) and Co3O4) have been effectively used to synthesize LiCoO2 by a simple thermal treatment. Different characterization techniques have proven the process to be efficient, simple and cost effective.
The work was conducted completely in the Heat Treatment Lab, Pilot Plant and Process Development Center, Bangladesh Council of Scientific and Industrial Research (BCSIR), Dhaka. The authors are greatly thankful to Dr. Abdul Gafur, PSO, IFRD, BCSIR and Dr. A. S. W. Kurny, Professor, Dept. of MME, BUET for their extended help. Dept. of GCE, BUET and BCSIR Laboratories, Dhaka, IFRD, INARS and BTRI, BCSIR also aided with analytical support.