The Effect of Shredding and Particle Size in Physical

Materials Science Forum Vols. 730-732 (2013) pp 653-658 © (2013) Trans Tech Publications, Switzerland doi:10.4028/www.scientific.net/MSF.730-732.653

The Effect of Shredding and Particle Size in Physical and Chemical Processing of Printed Circuit Boards Waste P.C. Oliveira1, a, F. Charters Taborda2, C.A. Nogueira1, b, F. Margarido2, c 1

LNEG – Laboratório Nacional de Energia e Geologia, UPCS, Estrada do Paço do Lumiar, 1649-038 Lisboa, Portugal 2

Instituto Superior Técnico, Technical University of Lisbon (TULisbon), Av. Rovisco Pais, 1049-001 Lisboa, Portugal a

[email protected], b [email protected] (corresponding author), c

[email protected]

Keywords: Printed circuit boards; Recycling; Shredding; Leaching.

Abstract. Circuit boards present in most electric and electronic devices are very important components, which should be removed during sorting and dismantling operations in order to allow further adequate treatment for recovering valuable metals such as copper, nickel, zinc, lead, tin and rare elements. This recovery can be made by physical and chemical processes being size reduction by shredding the first step. In this paper, the effect of particle size in physical and chemical processing of printed circuit boards is presented and discussed. Shredding using cutting-based equipment allowed the comminution of boards and the liberation of particles composed by different materials (mainly metals and resin). Particle sizes less than 1 mm seems to be appropriate to attain high liberation of materials, which is crucial for the physical separation using gravity or electrostatic processes. Concerning chemical treatment, hydrometallurgical processing involves a leaching operation which can be also influenced by particle size of shredded boards. Samples with different granulometries were leached with 1 M HNO3 solutions, being leaching yields evaluated. It was concluded that particle size can be an important factor for the solubilization of some metals, but the effect is not similar for all elements. When average diameters change from 2.0 to 0.20 mm, nickel, aluminium and tin reactivity were not significantly affected, being this effect important for copper. Zinc behavior was very dependent from extreme particle sizes but was less affected in intermediate granulometries. Lead leaching showed also a peculiar behavior, exhibiting high and almost constant yields (8090%) for particle size of solids up to 1.2 mm, and decreasing suddenly for higher granulometries. The effect of time on chemical reactivity for samples with different granulometries demonstrated that particle size affects reaction rates but eventually similar efficiencies can be obtained for long time periods. Therefore the relationship between results from shredding operation and chemical leaching step needs to be optimized, considering the balance between factors like consumption of energy during grinding operation and residence time in leaching. 1. Introduction Waste of electric and electronic equipment (WEEE) is one of the most important waste streams which correct management through material valorization is fundamental. In Europe (EU27), about 7 - 9 million tonnes of electric and electronic scrap are generated every year, corresponding to about 20 kg per year and per habitant. Printed circuit boards (PCB) are important components in all electronic devices and in most electric equipments, being part of the list of components which need to be removed from any collected WEEE, and sent to selective treatment, according with the European legislation. The importance of PCB is related with environmental and economic issues. Some substances in PCB can be potentially hazardous (e.g. heavy metals, halogenated organics), but metal contents are interesting in the economic point of view. PCB can be considered high grade copper concentrates (typically with 20All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of TTP, www.ttp.net. (ID: 85.243.107.176-16/08/12,22:44:35)

654

Advanced Materials Forum VI

30% Cu) containing also secondary metals such as zinc, tin, lead, nickel, as well as noble and rare metals. Therefore the management of these wastes through materials recycling will be the best approach. The most used route for recycling PCB is by pyrometallurgy using secondary copper smelters. However the use of ovens is only possible for high plant capacities. Hydrometallurgy treatment can provide more versatile alternatives for copper bearing materials. Some studies have already been carried out by researchers, using several leachants and conditions, namely using acids and oxidants. Nitric acid showed good possibilities for recovering copper and other metals from PCB [1,2]. Sulphuric acid alone does not allow achieving high metals recovery but its combination with hydrogen peroxide is a possible alternative already tested [3]. Leaching using mixtures of acids (H2SO4, HCl and HNO3) was also proposed, taking advantage of the simultaneous oxidizing and complexing effects of different acids [4]. Electro-assisted leaching was also studied by some authors, where external current was applied to promote oxidation, using H2SO4 and chloride media [5], or HCl media which generate chlorine [6]. A different option using thermal oxidation of metals followed by sulphuric acid leaching was also tested [7]. Finally, a bioleaching process for metals recovery from PCB was also published [8]. Results published on papers above described can be considered as a first step in the research on hydrometallurgical processing of PCB, since many other factors and conditions will be necessary to be tested to assess a comprehensive knowledge about all reactions, mechanisms and phenomena which can occur. Among others, an important factor in the leaching of PCB is particle size, but its effect has never been reported in literature. In the above papers, different solids characteristics have been used, varying from large fragments to fine shredded particles, so results and conclusions reported are hardly comparable. Solid-liquid reactions involve chemical transformations at the solid surface, therefore particle size can be crucial in the kinetics mechanisms and subsequently in the efficiency of metals recovery. In this paper, results of a study on the influence of shredding of PCB in the leaching with nitric acid solutions are presented and discussed. 2. Experimental Procedures PCB wastes used were obtained from motherboards of discarded computers. The boards were firstly cut in fragments with approximately 8x4 cm and then shredded in a cutting mill (Retsch SM2000). Several samples of PCB with different particle sizes were produced, namely very gross, gross, fine and very fine, by using different discharge grids. Samples with granulometry very gross and gross were obtained by shredding and using 6 and 4 mm grid apertures. Part of the very gross material was removed from the bulk and reground using a grid aperture of 0.75 mm and producing the fine material. Similar procedure was made for part of the gross fraction for preparing the very fine material, in this case using a 0.50 mm grid. The particle size distribution was determined by dry sieving (Retsch AS200) and the liberation of particles was assessed by stereoscopic microscopy (Nikon SMZ-2T). Solids used in the leaching trials were homogenised by sampling using a rotating divider. The leaching experiments were performed with materials previously described, in 200 mL capacity glass flasks and using 1 M HNO3 solution as leachant (prepared from reagent grade chemical), at liquid/solid ratio (L/S) of 25 L/kg, in a temperature control oven provided with orbital shaking (at 200 min-1 rotational speed). When necessary, aliquots of reacting pulps were periodically collected from each vessel, centrifuged and the liquor sent to metals analysis for evaluating the influence of time. In the end of each experiment, the solids and leaching solutions were separated by filtration, the solids being then washed with water and dried. Both solids and liquids were finally analysed to determine metals content and to evaluate the attained yields. The determination of metal contents in initial solids and leaching solutions was carried out by Atomic Absorption Spectrometry (AAS, GBC 906AA). Concerning solid samples, a preliminary dissolution step by acid digestion in microwave furnace (CEM MDS-2000) was used. After leaching, final residues were also analysed by X-ray Fluorescence (EDXRF, Spectrace QuanX) to validate the calculated leaching yields.

Materials Science Forum Vols. 730-732

655

3. Results and Discussion 3.1 Shredding and Characterization. The shredding of PCB wastes led to the production of four fractions whose particle size was assessed by the characteristic diameters showed in Table 1. For fine and gross fractions, the most used in this study, a detailed size distribution is presented in Fig. 1 (in the form of cumulative weight percent curves). Average diameters found were 2.1, 1.2, 0.55 and 0.20 mm, respectively for very gross, gross, fine and very fine samples. Microscopic observation showed that most of the particles with a size less than 1 mm were liberated, i.e. each particle is only constituted by one material. For very fine sample and for more of 95% of the weight of the fine sample, all particles presented these characteristics. Many copper-containing particles, resulting from grinded conductive tracks of PCB as well as from connecting pins, these ones usually present as copper alloys, were observed. Very clean particles of epoxy resin / fibber glass composite were also easily identified. By the contrary, in the coarse samples many multi-material particles were observed. In very gross sample, liberated particles were rare, and fragments of resin still containing copper tracks were common. In the gross sample, a mixture of liberated and unliberated materials was observed. From results obtained, it seems that efficient liberation occurs only for particle size less than 1 mm, these conditions being absolutely mandatory when physical separation operations are expected to be used. This is the case when operations for plastics separation by electrostatic, gravity or density methods, or for the separation of non-ferrous metals by “eddy current” technology are applied.

Fine, grid 0.75mm

Gross, grid 4mm

Cumulative finer (%)

100 80 60

Shredded sample Very gross Gross Fine Very fine

d10 (mm) 0.95 0.48 0.07 0.05

d50 (mm) 2.1 1.2 0.55 0.20

d90 (mm) 3.9 2.1 0.85 0.46

Table 2 – Chemical composition of PCB waste. Cu 29 27 31 30

Elemental composition (%) Zn Pb Ni Al Sn 5.3 4.2 0.21 5.1 7.5 5.1 3.7 0.22 5.6 4.0 5.5 3.1 0.35 7.2 5.0 6.2 3.2 0.40 7.7 4.7

Fe 0.45 0.60 3.80 1.40

Average

29

5.5

3.5

0.30

6.4

5.3

1.6

St.Dev.

1.5

0.5

0.5

0.10

1.3

1.5

1.6

Samples

40

(1) (2) (3) (4)

20 0 0.01

Table 1 – Characteristic diameters of shredded PCB.

0.1

1

10

Particle Size (mm)

Fig. 1 – Particle size distribution of two samples of shredded PCB. 3.2 Leaching of PCB. Concerning the chemical processing of PCB waste, the study focused the development of a hydrometallurgical recycling process, being the leaching with nitric acid the core operation. For copper, the main base metal in PCB, the following general reaction is expected to occur, (5-2x) Cu (s) + 2 NO3- (aq) + (12-4x) H+ (aq) = (5-2x) Cu2+ (aq) 2 NOx (g) + (6-2x) H2O (l) (1) where copper is oxidized to soluble copper ion by nitrate and generates NOx species, the most probable being NO or possibly N2O. In the leaching of metals the particle size play an important role, since it is required that solid particles react with the leaching agent, the reaction efficiency being thus depend on the following aspects: (1) the access of the acid to the metal particles; (2) the size of the particles. A substantial

656


fraction of the copper is present in the conductive track, inserted in the epoxy matrix, having at least a part of it on the surface. However other part of the tracks and pins are in the inner were the access is more difficult. It seems that shredding can be an important step, so can provide higher surface area, and higher exposure of the material to the chemical reaction. Nevertheless, the real effects of the particle size are not well known. It is also possible that the chemical agents such as H+ and NO3ions can migrate through the pores of the matrix to the solid surface, thus not requiring complete liberation of particles as obliged in physical separation processes. The effect of particle size was studied in a series of leaching experiments using as leachant 1 M HNO3 solution, for 3 hours at 40ºC. The leaching yields obtained for the main metals in PCB are presented in Fig. 2. It is observed that the leaching is improved for all studied metals, when particle size decreased. This effect is less evident for nickel, with the yields in the range 52-66%. For copper the effect is very pronounced, and the leaching efficiency changed from 5% in very gross sample to 58% in very fine sample. Very similar results were found for zinc. Concerning Al and Sn, the variation with granulometry was also observed but the leaching yields were always relatively low. The behaviour of lead was slightly different since the leaching efficiency was less pronounced in the three first samples (yields in the range 77-92%) and suddenly decreased for the very gross material to about 25%. This could be explained by the easier detachment of weld when compared with the other metal bearing components, except for the coarser fraction where partial occlusion of lead was also occurred.

Leaching Yields (%)

100 80

Pb

Zn

Ni

Al

Cu

Sn

60 40 20 0 0

0.5

1

1.5

2

Average Particle size (mm)

Fig. 2 – Influence of particle size in the leaching of shredded PCB (1 M HNO3, T=40ºC, t=3 h, L/S=25 L/kg). 3.3 Influence of particle size, time and temperature in metals leaching. The influence of particle size was also assessed for different reaction times and temperatures. The metals dissolution yields were evaluated at two levels of leaching temperature, 40 and 90ºC, using two PCB shredded samples, the fine and gross fractions. Fig. 3 shows the evolution of leaching yields of Cu, Zn, Ni and Pb as a function of time. For the other metals (Sn and Al) the results are not presented since the leaching yields achieved were always relatively low (below 5% and 15%, respectively). It is clear that particle size influences the leaching of the four metals, but with different importance. Lead is the less affected, however in the initial period of reaction the effect is still visible. For Cu, Zn and Ni, the influence of granulometry is observed in practically all range of time analysed. The figure also shows that temperature is a very important factor in what concerns leaching efficiency, but in the case of lead this effect was also less perceptible. The quantification of effects was statistically evaluated by factorial design methodology and the analysis of variance using Fisher distribution (Table 3). For these calculations, the factors particle size (d50), time and temperature were considered at two levels (high and low), respectively: 0.55 and

Materials Science Forum Vols. 730-732

657

1.2 mm; 1 and 4 h; 40 and 90ºC. Time and temperature are generally more significant than particle size. Only copper leaching is significantly affected (negatively) by the granulometry of the solids, zinc is in the limit of significance and for the other metals the confidence degrees were below 95%, therefore considered not significant. Although particle size affects leaching yields, the results obtained demonstrate that it is not absolutely crucial to achieve high efforts in grinding PCB, given that a reasonable average particle size is attained (e.g. d50 of about 1.2 mm) to allow the leaching of metals. 90ºC, fine material

40ºC, fine material



90ºC, gross material




100

100

(a) Cu

80

Zn Leaching (%)

Cu Leaching (%)

80

(b) Zn

60 40 20 0

60 40 20 0

0

1

2

3

4

0

1

Time (h)

3

4









100

80

80

Pb Leaching (%)

100

Ni Leaching (%)

2

Time (h)

60 40 20

60 40 20

(c) Ni

(d) Pb

0

0 0

1

2

3

0

4

1

Time (h)

2

3

4

Time (h)

Fig. 3 – Influence of time in the leaching of metals in two samples of shredded PCB (fine and gross), at two temperatures (40 and 90ºC); (L/S=25 L/kg, 1 M HNO3) Table 3 – Statistical analysis of effects of factors and respective significance. Effect on Leaching Yields

Factor

Statistic F

Confidence of Effect (%)

Cu

Zn

Ni

Pb

Cu

Zn

Ni

Pb

Cu

Zn

Ni

Pb

Particle size

-18.6

-6.2

-9.3

-3.3

61

10

7.9

3.2

99.6

95

93

83

Time Temperature

35

33

15.9

6.9

221

290

23

14

99.9

99.96

98

97

49

49

29

5.6

421

631

77

8.9

99.98

99.99

99.7

94

Note:

Analysis of variance was performed by comparing mean squares of effects with mean squares of standard errors estimated by 4 replicates at preset conditions.

658


4. Conclusions The shredding of PCB waste was efficiently performed using a cutting mill producing liberated single material particles below 1 mm size. Particle size affected chemical processing by acid leaching, but with different significance according with the metals. In the range of d50=0.55-1.2 mm, it was concluded that copper dissolution increased significantly when particle size decreases, while effect on zinc leaching was less pronounced, and for nickel and lead the effects were considered statistically insignificant. These results showed that shredding of PCB can be advantageous into certain limits but over-grinding is needless and shall be avoided in order to save energy and costs. Acknowledgement: The financial support made by Amb3E (Associação Portuguesa de Gestão de Resíduos de Equipamentos Eléctricos e Electrónicos) is gratefully acknowledged. References [1] T. Kinoshita, S. Akita, N. Kobayashi, S. Nii, F. Kawaizumi and, K. Takahashi, Metal recovery from non-mounted printed wiring boards via hydrometallurgical processing, Hydrometallurgy 69 (2003) 73–79 [2] A. Mecucci and K. Scott, Leaching and electrochemical recovery of copper, lead and tin from scrap printed circuit boards, J. Chem. Technol. Biotechnol. 77 (2002) 449-457 [3] C.J. Oh, S.O. Lee, H.S. Yang, T.J. Ha and M.J. Kim, Selective Leaching of Valuable Metals from Waste Printed Circuit Boards, J. Air & Waste Manage. Assoc. 53 (2003) 897–902 [4] L.A. Castro and A.H. Martins, Recovery of tin and copper by recycling of printed circuit boards from obsolete computers, Braz. J. Chem. Eng. 26(4) (2009) 649 - 657 [5] Z. Ping, F. ZeYun, L. Jie, L. Qiang, Q. GuangRen and Z. Ming, Enhancement of leaching copper by electro-oxidation from metal powders of waste printed circuit board, J. Hazard. Mater. 166 (2009) 746–750 [6] E.-Y. Kim, M.-S. Kim, J.-C. Lee, J. Jeong, B.D. Pandey, Leaching kinetics of copper from waste printed circuit boards by electro-generated chlorine in HCl solution, Hydrometallurgy 107 (2011) 124–132 [7] T. Havlik, D. Orac, M. Petranikova, A. Miskufova, F. Kukurugya and Z. Takacova, Leaching of copper and tin from used printed circuit boards after thermal treatment, J. Hazard. Mater. 183 (2010) 866-873 [8] J. Wang, J. Bai1, J. Xu and B. Liang, Bioleaching of metals from printed wire boards by Acidithiobacillus ferrooxidans and Acidithiobacillus thiooxidans and their mixture, J. Hazard. Mater. 172 (2009) 1100–1105