RECOVERY OF HEAVY METALS FROM SPENT NICKEL-CADMIUM BATTERIES

 

 

Lia I.C. Barros, Adriano M.G. Pacheco*, and Fernanda Margarido

 

CVRM - Mineral Resources Valorization Centre, Instituto Superior Técnico

Av. Rovisco Pais 1, 1049-001 Lisboa, Portugal (* mail to: apacheco@ist.utl.pt)

 

 

Abstract

 

Ni-Cd cells remain the main sink for Cd production and, at the same time, the major source of Cd pollution to the environment, which calls for an improvement in its recovery. In this study, dismantled batteries underwent granulometric separation by wet sieving, using an intense flow. Regardless of the shredder’s output, results point to a major separation around 2.8-mm mesh. Upper fractions still contain heavy metals though, therefore an additional process is required. The infra fractions can proceed directly to the recovery stages downstream. The amount of material in the finest fractions is inversely proportional to the mesh of the shredder's output, yet this parameter appears to have no influence on the granulometry of such fractions. Toxic metals (Cd, Co, Ni) concentrate in the infra fractions, whereas ferrous and organic scrap are retained above.

 

Introduction

 

There is not much doubt about the significant increase in collection rates of spent batteries in the European Union, following the EC Battery Directive 91/157 EEC (1991). The present situation is still far from being satisfactory though. In 1997, total EU take-back was up to 20,000 tons, yet almost 30,000 tons (about 900 million portable units) were being sold every year in the German market alone [2,3]. Related figures from other Member States point to a collection rate of about 50-60 % at best, an efficiency that seems unlikely to rise without some sort of economical incentive to end-users for proper battery disposal [1,4]. Rate notwithstanding, an important amount of toxic waste from spent cells will keep on getting into incinerators, landfills and other general-purpose, disposal sites in the (near) future. Now the problem for an average, small-to-medium country in a developed stage is that there are only two alternatives: the former one, which means no chemical spillage – just in the strict sense, of course…! – but also no metal recovery whatsoever, or else feeding the spent cells into an undifferentiated, scrap battery (or scrap metal) processing plant, which could mean some recovery yet involve some costly and probably off-border operation, given the scarcity of specialised reclamation facilities around the world.

The above considerations triggered the concept behind a current project on metal recovery from spent Ni-Cd batteries, which this paper is about. That and also an idea of putting together an all-inclusive operation, from scrap batteries to value metals, which could avoid high temperature processing and huge plants, and rather fit into small-to-medium scale units without loss of profitability. The Project runs on a multi-partnership basis, involving people and resources from a technical university, a national laboratory and a private corporation, with financial support from the Portuguese Government and the European Union. Major research tasks are:

i.      Physical processing of spent batteries, in order to optimize the granulometry and the recyclable content downstream.

ii.     Hydro-electrometallurgical processing of battery scrap into nickel, cadmium and cobalt forms, as pure as possible.

iii.    Assembly of unit operations into a laboratory-scale prototype for the whole metal-recovery process.

iv.   Technical and economic audits.

The present paper will focus on the physical-processing stage. The separation techniques as well as the characterization of the resulting fractions are presented and discussed, especially in what matters to the easing of subsequent operations.

 

Methodology

 

The study herein deals with the initial (physical) stage of an integrated process for recovering Cd, Ni and Co from spent Ni-Cd batteries. All experiments refer to small batteries of the cylindrical, sealed type, picked up at random from batches of fully-exhausted devices that had been used in campaign radios by the Portuguese Army. Each unit weighs 150±5 g and has a bright-yellow, plastic cover. Cells were dismantled by means of a single-shaft disintegrator (ERDWICHâ; model: EWZ 200), with interchangeable output sieves of 20, 10 or 6 mm. The resulting material underwent granulometric separation by wet sieving (down to 0.355 mm), using an intense water flow (10 l min-1) and FRITSCHâ sieves (DIN 4188 series) of 1.4, 2.0, 2.8, 4.0, 5.6 e 8.0 mm.
The samples mass was determined through the Pierre Gy Theory [8].The granulometry of the fractions under 0.355 mm, resulting from wet sieving with moderate flow experiments, were checked out by laser diffraction analysis, using a CILASä 920 analyzer. Samples dismantled with the three output sieves were analyzed. Optical and scanning electron microscopy with an energy-dispersive probe (SEM/EDS) were used for morphological and chemical characterization of all fractions.

 


Results and Discussion

 

Cell Dismantling

 

This operation promotes a first splitting between the value metallic material (Cd, Ni and Co) and the scrap (metallic case, plastic wrap and the electrode material supporting grates). Two fractions are obtained – the upper and the infra. There are no significant differences between them, only the upper one usually comes richer in plastic cover and metallic case. In an industrial process, however, some advantage could be taken of handling the two fractions separately. The volume of dismantled material decreases proportionally to the mesh of the shredder's output sieve. The hardness of the material, as well as the cutting process itself, leads to a random distribution of components in either fraction, which makes it difficult to work out an association (if any…) between material losses and such an output.

 

Granulometric Classification

 

Results from wet sieving and laser granulometry (< 0.355 mm) are summarized in Figures 1 and 2,  respectively.

 

 

 

 

 

 

 

 

 

 


Fig. 2 - Simple histograms of the infra and upper fraction distribution as a function of the used separation sieve.

 

Up to 2.8 mm sieve included, the upper fraction amount is always bigger than the infra one, whatever output shredder’s sieve is used. From the 5.6 mm up to 10 mm sieve, this situation is inverted only for the material dismantled with the smaller mesh shredder’s output sieve (6 mm). In all wet-sieving experiments, the mass difference between upper and infra fraction diminishes with the dimension of the shredder’s output sieve mesh. In the assays using the 1.4 , 2 and 2.8 mm sieves, a very little amount of plastic and practically no metallic case appears in the infra fraction. Infra 4 mm fraction already contains some grate but, still thus, its composition does not differ significantly from the previous fraction (infra 2.8 mm). Anyway, the plastic is not problematic, therefore its splitting is relatively simple, since is the only not magnetic component of the batteries. For 5.6 , 8 and, mainly, 10 mm sieves, the amount of grate and metallic case in the infra fraction is very high and the splitting between the components moves away, each time more, from the desired one. This situation occurs, probably, because these values are similar to the shredder’s output sieves dimensions, despite the difference in mesh geometry. Thus, based in the obtained results, it can be concluded that, for wet-sieving operation using the assayed water flow, the cut-off value rounds 2.8 mm or, at the most, 4 mm.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


Fig. 2 – Cumulative plots of grain size and population density distribution for the wet sieving finer fraction.

 

On the basis of the obtained results it may be concluded that shredder’s output sieve dimension does not influence the finest material fractions granulometry, but only its amount, which is inversely proportional to the sieve’s mesh dimension. Being the main objective to obtain the biggest amount of individualized metallic dust of the remainder material, the use of 6 mm shredder’s is far more adjusted. However, in this case, the plastic and the metallic case appear, especially in the upper fraction, in smal ler particles, witch can turn its recovery far more difficult.

 

Morphological and Chemical Characterization

 

After analyzing the different batteries components, we find plastics and metallic crap; value metals reduced to dust and their respective supporting grate constitute the electrodes. A different coloration between the anode (clear grey) and the cathode (dark grey) is visible. Both materials present a strong tack to the supporting grate, specially the anodic one, making more difficult its full splitting. Figure 3 presents microphotographs obtained in the Scanning Electron Microscope, evidencing the most common morphologies of the anodic and cathodic material. These samples were subjected to mechanical dismantling, followed by wet-sieving with moderate water flow [6]. The X-ray spectrum obtained by Energy Dispersive Spectrometry (EDS), for details of anodic and cathodic material, are showed in Figure 4.


                                                              
a)                                                                                                     b)

Fig. 3 – Scanning Electron Microscope microphotograph of the electrode material:

a)       anodic (2000x); b) cathodic (1000x).

 


a)                                                                                                                       b)

 


Fig. 4  – Electrode material X-ray spectra, obtained by Energy Dispersive Spectrometry:

a) anodic; b) cathodic.

 

The anodic and cathodic material appears in the samples with varied morphologies, but it is not very difficult to distinguish them: the anodic material evidences a more compact and geometric structure that the cathodic one. Then it is comprehensive that this offers the biggest resistance to the splitting from the respective supporting grate. The spectra analysis enables to verify that the anodic material grains are essentially constituted by cadmium. The supporting grate is constituted by iron and nickel. The cathodic material is essentially composed by nickel, some cadmium and a small amount of cobalt.

Conclusions

 

All experiments herein refer to single-sieve wet sieving tests, using an intense water flow - 10 Lmin-1 - during 2 minutes. Another experiments already described [6] were done with different experimental conditions: using a sieve serie and a 1Lmin-1 washing water flow for over 30 minutes In the first experiments the separation between scrap (iron and organic material) and the value metals (Cd, Co, Ni) occurred, mainly, at the 1.4 mm sieve level, despite of the shredder’s output sieve mesh dimension. The use of an intense water flow made possible to increase the cut-off value for 2.8 mm, or even 4 mm, which constitutes, from the technical-economic point of view, a more viable solution. According to this experiments, it can be verified that the intense water flow wet sieving operation is the most adjusted after the dismantling, allowing to clean the scrap (plastic and metallic case) from the metallic dust. Therefore, a magnetic process (for example) can easily do these components removal. It was not possible to achieve full splitting of the electrode material from the respective supporting grate, even when an intense water flow was used. Thus, it will be necessary to apply for chemical processes. Leaching assays will be able to constitute an important contribution to the splitting process. Regarding to the organic fiber, is separation from electrode metals seems fundamental, since it can turn impracticable their reuse. In this phase, the definitive choice of the optimal shredder’s output sieve mesh, as well as a definitive wet-sieving cut-off value, can not be put to stray. The final conclusions can only be made when subsequent stages of grain size analysis and, over all, chemical (leaching and solvent extraction) and electrochemical splitting experiments come to an end.

 

Acknowledgements

 

The work, which this paper refers to, has been financially supported by the Foundation for Science and Technology (FCT; Portugal) under the contract PRAXIS XXI 3/3.1/CEG/2574/95. Lia Barros acknowledges a research scholarship (PRAXIS XXI/BIC/17099/98) from FCT as well. The Authors are indebted to the Portuguese Army (DGMT) for providing the raw material (spent batteries) for this continuing study.

 

References

 

1.     Rahbek LW (1998), In: Proc. Fourth International Battery Recycling Congress (Hamburg, Germany; July 1-3, 1998), Paper # 6.

2.     Oshitani M, Takayama T, Takashima K, Tsuji S (1986), J. Appl. Electrochem. 16: 403-412.

3.     Beck M (1998), Recycling International 1(4): 18-23.

4.     Fricke J (1998), In: Proc. Fourth International Battery Recycling Congress (Hamburg, Germany; July 1-3, 1998), Paper # 4.

5.     Salhofer S (1998), In: Proc. Fourth International Battery Recycling Congress (Hamburg, Germany; July 1-3, 1998), Paper # 5.

6.     Barros LIC, Guimarães C, Margarido F, Pacheco AMG (1999), In: Proc. Fourteenth Annual Battery Conference on Applications and Advances (Long Beach CA, USA; January 12-15, 1999; HA Frank, ET Seo, Editors), Piscataway NJ (USA), IEEE, pp. 297-302.

7.     Van Deelen CL, Van Erkel J (1990), In: Nickel-Cadmium Battery Update (Brussels, Belgium), London (UK), Cadmium Association, pp. 72-75.

8.     Gy P, Théorie Générale de l’Echantillonnage des Minerais en Vrac, St. Etienne, Société de l’Industrie Minérale.