froth flotation for beneficiation of printed circuit boards ...

This article was downloaded by: [Iyiola Otunniyi] On: 25 April 2013, At: 09:24 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Mineral Processing and Extractive Metallurgy Review: An International Journal Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/gmpr20

FROTH FLOTATION FOR BENEFICIATION OF PRINTED CIRCUIT BOARDS COMMINUTION FINES: AN OVERVIEW a

I. O. Ogunniyi & M. K. G. Vermaak

a

a

Materials Science and Metallurgical Engineering, University of Pretoria, Pretoria, South Africa Version of record first published: 30 Mar 2009.

To cite this article: I. O. Ogunniyi & M. K. G. Vermaak (2009): FROTH FLOTATION FOR BENEFICIATION OF PRINTED CIRCUIT BOARDS COMMINUTION FINES: AN OVERVIEW, Mineral Processing and Extractive Metallurgy Review: An International Journal, 30:2, 101-121 To link to this article: http://dx.doi.org/10.1080/08827500802333123

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/termsand-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be

Downloaded by [Iyiola Otunniyi] at 09:24 25 April 2013

independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

Mineral Processing & Extractive Metall. Rev., 30: 101–121, 2009 Copyright Q Taylor & Francis Group, LLC ISSN: 0882-7508 print=1547-7401 online DOI: 10.1080/08827500802333123


FROTH FLOTATION FOR BENEFICIATION OF PRINTED CIRCUIT BOARDS COMMINUTION FINES: AN OVERVIEW

I. O. OGUNNIYI AND M. K. G. VERMAAK Materials Science and Metallurgical Engineering, University of Pretoria, Pretoria, South Africa

End-of-life printed circuit boards (PCBs) are a complex secondary resource stock that continues to evolve more complexly in new generation devices. This is a major challenge in its physical processing. The value of precious and base metals loss to the 75 mm fraction generated during PCB comminution has been identified as a major constraint toward improving physical processing. Application of froth flotation for beneficiation of this fines fraction is conceivable, as each group of particles in the mix has distinct surface properties that can be exploited. However, recognizing the complex flotation system such a sample can produce, an informed discourse has been presented herein: starting from a characterization of the resource stock, then to a review of almost three decades of PCB physical processing, before attempting to simplify the flotation complexities. This overview is intended to be a springboard for detailed investigations into this theme. Keywords: comminution fines (CFs), end-of-life printed circuit board (PCB), froth flotation application, physical processing, resource characterization

Address correspondence to I. O. Ogunniyi, Materials Science and Metallurgical Engineering, University of Pretoria, Pretoria 0002, South Africa. E-mail: mrolatunji@tuks. co.za

102

I. O. OGUNNIYI AND M. K. G. VERMAAK


1. INTRODUCTION Physical (or mechanical) processing is recognized as the most environmentally friendly alternative for resource recovery from end-of-life printed circuit board (eol PCB) (Goosey and Kellner 2002). Precious and base metals value loss to the 75 mm comminution fine (CF) has been identified as a major constraint to possible extensive physical processing of eol PCB (Goosey and Kelner 2002). In a detailed overview toward improving PCB physical processing, it has been noted that froth flotation can be an effective beneficiation technique for value recovery from these fines (Ogunniyi and Vermaak 2007). Selectivity as well as handling 75 mm sizes are known merits of froth flotation techniques in the processing of complex and finely liberated ores. Froth flotation has also been applied in processing municipal solid waste (Shent et al. 1999; Dodbiba et al., 2002; Alter 2005). PCB fines consist of a mix of particles of metals, alloys, ceramics, and plastics, each with distinct surface properties that should enable selective wetting and make froth flotation separation possible. However, the material mixture in this context is fairly complex. Table 1 gives an indication of possible target values that can be expected in eol PCB CF: more than 10 base and precious metals, further complicated by diverse ceramics and plastics particles. Recognizing that natural mineral deposits with three or four target values are considered to be a complex orebody (Wills 1997), it is easy to appreciate that the beneficiation of the PCB CF via froth flotation can be quite involving. Thus, this article discusses pertinent issues in this prospect and can be a springboard for investigations on the theme. Broadly, two core issues are discussed: the logical flotation schemes that can be considered and the candidate surfactants that can be employed. In this connection, characterisation of PCB, and a brief review of the physical processing trend, are considered to be necessary background and will be addressed first. 2. CHARACTERIZATION OF EOL PCB In conventional minerals processing, a thorough understanding of the resource deposit to be exploited is the primary step. Hence, the following characterization, which attempts to present the occurrence, reserve, and representative compositions of eol PCB stream as a type of ore, while the physical processing can also be described as an applied mineral-processing problem.

FROTH FLOTATION FOR PCB COMMINUTION FINES

103

Table 1. Representative material compositions of PCBs in weight percent


Materials

%1

%2

%3

% 4

%5

%6

%7

Metals Cu 20 26.8 10 15.6 22 17.85 23.47 (Max.40%)1 Al 2 4.7 7 — — 4.78 1.33 Pb 2 — 1.2 1.35 1.55 4.19 0.99 Zn 1 1.5 1.6 0.16 — 2.17 1.51 Ni 2 0.47 0.85 0.28 0.32 1.63 2.35 Fe 8 5.3 — 1.4 3.6 2.0 1.22 Sn 4 1.0 — 3.24 2.6 5.28 1.54 Sb 0.4 0.06 — — — — — Au=ppm 1000 80 280 420 350 350 570 Pt=ppm — — — — — 4.6 30 Ag=ppm 2000 3300 110 1240 — 1300 3301 Pd=ppm 50 — — 10 — 250 294 SiO2 15 15 — 41.86 30 — — Ceramic 6 — — 6.97 — (Max 30%)1 Al2O3 Alkaline and 6 — — CaO 9.95 Alkaline MgO 0.48 earth oxides Titanates, Mica, etc 3 — — — — — — Plastics Polyethylene 9.9 — — — 16 — — (Max 30%)1 Polypropylene 4.8 Polyesters 4.8 Epoxies 4.8 Polyvinyl2.4 Chloride Polytetra2.4 Flouroethane Nylon 0.9 1

Shuey et al. 2006 from Sum, 1991; 2Zhao et al. 2004 from Lehner 1998; 3Cui 2005 from Zhang and Forssberg 1997; 4Kim et al. 2004, Incinerated PCB Product; 5Iji and Yokoyama 1997; 6Kogan 2006; 7ICP-OES Analysis of cellular phone PCBs with hot aqua regia digestion.

PCB mechanically supports and electrically connects electronic components using conductive pathways, or traces, etched from copper sheets laminated onto a nonconductive substrate (Wikipedia: PCB). Alternative names are printed wiring board (PWB) and etched wiring board (EWB). Populated PCBs have mounted components, while unpopulated PCB is only reinforced resin and printed copper wiring laminate. The structure has changed from very early one-sided boards to about 16 or 20 layered laminates in more recent high-tech boards, and as many as 50 in boards for special application (Holm 2001; Coombs 2001). PCB


104


is found in almost all electronics applications, spanning personal computers, telecommunication equipment, hospital equipment, cellular phones, and various consumer electronics. Many material compositions of the PCB have been reported, some of which are presented in Table 1. The data show the PCB as a respectable polymetallic resource, assaying far above many precious and base metals natural deposits. If ores are classified as such basically for the valuable contents, eol PCB can be rightly described as a type of ore=industrial ores (Castro et al. 2005). Unlike naturally occurring ores containing minerals values with definite occurrences and known deposits, eol PCB is spread over the globe. The occurrence of values and its ‘‘mineralogy’’ are functions of continuing technological innovations and changing legistrations. This translates into a high rate of obsolescence of electronics equipment, with increasing tonnages of eol PCB in waste electrical and electronics equipment streams. Figures of accumulations from countries have been compiled. In the United States, for personal computers (PCs) alone, 63 million units were estimated to be obsolete in 2005 (Kang and Schoenung 2005). In the United Kingdom, about 50 000 tons=annum of waste PCBs is generated, with about 40 000 tons as populated PCBs and 10 000 tons as unpopulated or associated boards manufacturing scrap (Goosey and Kellner 2003). The estimate for 2006 was 100500 tons (Shuey et al. 2006). A review written in 2000 in Taiwan claims about 300 000 scrap PC units per annum (Lee et al. 2000), which had risen to 700 000 by 2004 (Lee et al. 2004). With generally few statistics for South Africa, Desco Electronic Recyclers was reported as processing 400 tons of PCB per annum (Furter 2004). Another report quotes more than 500 000 PCs dumped per year (Mackay 2004). Intel South Africa, for instance, has been in business for more than 35 years and had sold in excess of 1 billion CPUs, implying that there are many obsolete PCs dumped (Mackay 2004). Large amounts of undocumented accumulation and scatters of PCB occur around the world, many due to shipment of used PC to developing countries. A 2005 report claims that about 400 000 used computers were imported to Nigeria monthly, almost 75% of which are neither usable nor economically repairable or resalable (Laurie 2005). This situation is typical of many developing countries (Laurie 2005). At about 2.75% by weight of PCB per PC unit, and one PC unit with monitor weighing about 27.5 kg (Shuey et al. 2006), along with Table 1, the huge tonnages of metallic, plastic, and ceramic material resources in eol PCB can be appreciated.



105

These figures show a large resource stream with many values that are very inviting to be exploited, but a consequence of this diverse valued constituent is the challenge of beneficiation into clean material fractions. Chalcopyrite (CuFeS2), for instance, is the main ore mineral of copper and the best copper assay from its processing is 34.7%. This concentrate is amenable to very straightforward matte smelting chemistry, oxidizing iron sulphide to the slag phase and sulphur to the gas phase, leaving metallic (blister) copper (Rosenqvist 1988). In contrast, in PCB nonferrous electrostatic separation products, at an assay of, for instance, 90% copper, the 10% complementary matter is not from a single mineral element (such as sulphur alone with copper in chalcocite) or a few definite co-occurring elements, but many contaminants to make the smelting and refining processes quite extensive. While ores containing three target minerals are considered to be complex in classical minerals processing, with extensive beneficiation operations, common eol PCB beneficiation flowsheets are generally too brief for the task. Improving eol PCB beneficiation demands recognition of the material complexity involved, as well as the development of effective beneficiation operation to match. 3. PHYSICAL PROCESSING TREND OF EOL PCB Physical processing of PCB entails a series of stage-wise comminution– separation operations, separating metallic and nonmetallic values into different materials fractions. Various separation technologies are employed in handling the comminution products at different stages. These include magnetic separation, eddy current separation, electrostatics, air table, gravity air classifier, air cyclones, and screens, among others (Goosey and Kellner 2003; hamos GmbH; Kang and Schoenung 2005; Li et al. 2004). The comminution and separation techniques have continued to change in the almost three decades of these operations. Figures 1 and 2 compare typical physical processing flow sheets in about a 20-year interval (Lee et al. 2004; Kravchenko et al. 1983). From early practices (Figure 1), differential dismantling is the first stage before pulverizing the remaining (evacuated) boards, and this matured to automated disassembly (Feldmann and Scheller 1995; Iji and Yokoyama 1997; Kravchenko et al. 1983; Li et al. 2004). Automating PCB disassembly is onerous—the diversity of boards, different joineries, etc.—the issues were many and many efforts were concerted


106


Figure 1. Processing of multicomponent REE scrap (Kravchenko et al. 1983).

on this (Stennett and Whalley 1999; Feldmann and Scheller 1995). Recovered components are refurbished and reused. Advancement in production and materials technology makes components quickly obsolete, not reusable, and new ones much cheaper. This has dowsed much of the drive for automated disassembly. With this trend, much of the emphasis for easy disassembly design considerations, at least with respect to PCB, could really become passe´ (Taylor 2002). Although efforts toward improving disassembly continue (Elif and Surendra 2006), comminution of the whole populated boards in low-speed, high-torque shear shredders is now more typical, with preremoval only of hazardous components such as batteries, capacitors, LCDs, etc. (Lee et al. 2004). Ferrohydrostatic separation (FHS), also called magnetohydrostatic separation, was applied to separate plastics, aluminum, and a heavier nonferrous mixture (Figure 1) (Kravchenko et al. 1983). This technology employs ferrofluids (colloidally stable and magnetizable fluids) to produce a medium whose apparent density can be varied up to a specific



107

Figure 2. Scrap PCB physical separation flowsheet, Huei-Chia-Dien Company, Taiwan (Lee et al. 2004).

gravity above 20 in fine steps, making it possible to achieve sink–float separation to tenths of a specific gravity. FHS had earlier been applied for municipal solid waste (Khalafalla 1973, 1976; Vesilind and Rimer 1981; Zhan and Shenton 1980). Over time, electrostatic separation took over this application of FHS in eol PCB physical processing (Gold and Dietz 2003; Iji and Yokoyama 1997; Li et al. 2004). Although a report comparing FHS to electrostatics separation is rare, the general preference for dry operation can be assumed to justify the shift. Considering the high density based separation efficiency that FHS promises, investigations comparing it to electrostatic separation on a broad basis (electricity consumption, throughputs, separation efficiency, and effective feed particles size range) need to be considered. A trend reversal that can improve PCB physical separation may result. At coarser sizes for electrostatics separation, eddy current separator (ECS) was later employed. ECS is ideal for selecting high-conducting, low-density fragments at sizes from about 50 mm down to 5 mm. With


108


improvement in permanent and electromagnetic rotors, there have been some developments to handle smaller feed sizes. Wet ECS and angular drum ECS have evolved from the horizontal drum ECS, with claims of good separation down to 2 mm (Cui 2005; Lungu 2005). However, between the lower size limit of ECS and the upper limit of electrostatics, separation remains a window allowing poor separation efficiencies and contaminated products for which flowsheets must carefully account. Some gravity air classifications and air cyclones can close up this gap, while opening chances of value losses. Generally, the trend has shown a preference for dry operations, probably to avoid the cost of process water and effluent treatments. This has been a general preference in waste processing (Kellerwessel 1993). With the complexity of PCB stock, the selectivity achievable with wet processing may be indispensable in achieving cleaner fractions. Preventing fine particles from becoming airborne can also cut losses and increases recovery. Many wet concentration techniques are used in minerals processing. Dense medium separation investigation has been reported for þ1.7 mm–4.75 mm PCB comminution products (Zhang and Forssberg 1997b) and there are improved bidirectional dense medium drums (ESR 1994). PCB physical processing may need to look more into wet density-based separation at industrial levels. However, at fine sizes the particle–fluid dynamics make simple sink–float separation less effective. The fine fraction really demands more attention. 4. COMMINUTION FINES Different stages of comminution are indispensable for the liberation of the PCB values. Size distribution of the comminution products gives significant proportion of the 75 mm fraction (Zhao et al. 2004; Kogan 2006). Up to 26% was reported in 1-mm closed grinding of the boards (Table 2; Zhao et al. 2004). The entire submillimeter particles go to electrostatics separation, whose efficiency is known to be poor at 100 mm (Kelly and Spottiswood 1989). The grinding products of Zhao et al. (2004), given in Table 2, were used in a comparative investigation of corona electrostatic separation and a type of air classifier referred to as a pneumatic separator. For the electrostatic separation, values for 75 mm were poor and not reported, while the pneumatic separation could only give copper recovery and grades of 27.83 and 26.91, respectively (Table 3). Even in the size ranges where the separation efficiency


109

Table 2. PCB grinding size distribution (extracted from Zhao et al. 2004)


Wt. %

Size range mm

1-mm closed-circuit hammer mill grinding

Open-circuit hammer mill grinding

þ1 1 þ 0.5 0.5 þ 0.25 0.25 þ 0.125 0.125 þ 0.075 0.075

0 40.87 11.73 15.23 5.54 26.73

15.86 35.26 10.10 12.86 4.70 21.22

can be acceptable, a broad resource-recovery consideration shows that the value loss is still significant. Yokoyama and Iji (1997) showed that the glass fiber and epoxy resin nonconducting fraction from the 300 to 100 mm size range contain 2.1% copper that is not recovered. This assay shows the waste fraction to be richer than many economically mined copper porphyry ores. The loss of precious metals is another issue. On the edge connectors, gold is plated just to a few microns thick. Any abrasion during grinding liberates them into this fines range, with the possibility of them being unrecovered. The cyclone dust fraction in a comminution operation was calculated to contain about 51% of the gold content of the feed (Kogan 2006) and up to 10% of the loss in this value is possible in physical processing (Goosey and Kellner 2002). Possible exits are separation inefficiencies or the fines particles becoming airborne. Table 3. Copper recovery and grade versus size range and separation technology (extracted from Zhao et al. 2004). Pneumatic Separation (Air classifier)

Corona electrostatic separation

Size range, mm

Recovery, %

Grade, %

Recovery, %

Grade, %

1.0 þ 0.5 0.5 þ 0.25 0.25 þ 0.125 0.125 þ 0.075 0.075

70.11 66.42 90.76 80.35 27.83

49.08 54.66 49.54 29.98 26.81

97.65 97.88 43.78 37.57 —

53.75 71.61 90.62 90.90 —


110


This shows that the fines fraction, about 75 mm in size, generated during comminution impacts the achievable metallic value recovery of the operation. To avoid these losses, some operations prefer to avoid physical processing as much as possible and incinerate the whole feed stream (Goosey and Kellner 2002), despite the off-gas handling problems or the hazard of releasing toxic furans, dioxins, and noxious gases (D’Silva 2004; Nourreddine et al. 1998). The right approach should be the investigation and adoption of comminution schemes that optimize liberation with minimal fines generation, as well as seeking alternative beneficiation techniques for the value in the minimized fines. An effective solution to the comminution fines problem can be a major step forward in PCB physical processing. Toward fines minimization, Iji and Yokoyama (1997), using a table and roller pulverizing mill (effectively a high-pressure grind roll, HPGR), showed that copper traces liberate in a coarser size range compared to the resin and plastics in the evacuated board laminates. Normally, because metallic materials are ductile, they will not shatter into fines in the event of comminution, and the fact that the mill employs compression and shear could also leave the copper traces with less breakage compared to glass fibers and plastics. On the other hand, the swing hammer mill that appears to have become the industry standard for eol PCB grinding (Goosey and Kellner 2003; hamos GmbH; Sander et al. 2004), breaks mostly by heavy impact, with abrasion shear and compression. This is a more complex stressing action compared to the HPGR, but populated eol PCB may as well require such for effective liberation. An investigation comparing these two comminution equipments is rare. The possibility of fine reduction with HPGR may justify trying it on populated eol PCB. The spherizing effect in the hammer mill, which reduces ECS efficiency (Zhang and Forssberg 1999), may be another reason. Increasing the number of stages of comminution–separation is another possible way to reducing fines; separating immediately after liberation is achieved, minimizing further breakage of the liberated pieces. In the quest for an effective beneficiation technique, froth flotation appears quite promising. Selectivity as well as handling relatively fine sizes are known merits of the froth flotation technique in processing complex and finely liberated ores. The selectivity has also been exploited in municipal solid waste processing (Alter 2005; Dodbiba et al. 2002; Shent et al. 1999). PCB fines consist of a mixture of particles of metals, alloys, ceramics, and plastics with distinct surface properties that should


111

enable selective wetting and make froth flotation separation possible. However, the mixture is fairly complex and will be an interesting challenge to the versatility of froth flotation technique. In this connection, two primary questions will be what logical froth flotation schemes could be applied, and with what type of reagents.


5. FROTH FLOATATION SCHEMES The froth flotation pulp, before aeration, is ideally a mixture of only two types of particles: hydrophobics and hydrophilics. Surface hydrophobicity can be naturally inherent or imparted by reagent conditioning. The system may rather demand depression of some particles while others retain their floatability. PCB comminution fines will represent a fairly complex flotation system; the following are considered as probable schemes that can simplify the operation. 5.1. Natural Hydrophobic Response Getting the particles into the two basic groups of hydrophobics and hydrophilics is achieved either by reagent conditioning or by exploiting natural surface property differences in the mixture, as it obtains in coal floatation. For PCB CF flotation, this can be a very useful scheme to start with. The particles contained in the fines being so diverse, some should exhibit natural hydrophobicity, e.g., particles from comminution of plastic components. Such particles are expected to float under their natural hydrophobicity, thus making it possible to achieve an initial upgrade of the bulk. This can also be considered to be a kind of reverse flotation in respect to the metallic values, before conditioning the resulting streams for more selective floats. 5.2. Gamma Depression Should significant natural hydrophobicity exist in the fines sample, an inherent problem then is entrainment. The sample contains fibrous particles of copper traces and glass fiber, which under heavy mass pull can entrain easily. Also, natural hydrophobicity can impede the wetting of fines and make pulping difficult, as small water-tight globules lock up and shield hydrophilic particles. Such globules covered with hydrophobic particles can easily report to float under a natural hydrophobic response. The introduction of depressant to achieve better wetting, easier


112


pulping, and selective depression of some otherwise froth-phase–bound particles can be a way to resolve this. For such a complex pulp, with plastic particles among possible targets for depression, chemical depressants will not be ideal. Rather, the concept of gamma flotation can be employed. Gamma flotation is based on the critical surface tension (cC) of wetting of solids (Buchan and Yarar 1995), similar to the original Zisman’s critical surface tension concept (Zisman 1965) described by Alter (2005). This entails selective flotation by selective wetting through reducing surface tension of the pulp below that of some otherwise hydrophobic particles, which become wetted. As employed in plastic flotation, the surface tension of water (cL=V) is reduced with surfactants such that it lies between the critical surface tensions (cC1 and cC2) of two plastics or groups of plastics: cC1 < cL=V < cC2 : Plastic particles with cc cC2 thus become wet and report to the sink fraction. In this scheme for PCB CF flotation, the reduction in surface tension (c) will serve to achieve depression and has been preferably described as the gamma (c) depression. 5.3. Bulk Metallic Flotation Another probable simplification approach to this flotation system is bulk metallic flotation. Metallic surfaces, with generally higher surface energies than those of plastics, can be expected to react with collector molecules (chemisorption) forming hydrophobic surfaces (Woods 1996). This can be exploited for a bulk flotation of metallic particles in the mix and is comparable to the conventional flotation of ores such as the galena– sphalerite–chalcopyrite complex, where all the metal value-bearing minerals can be floated first from co-occurring gangues and the bulk float treated in some selective order (Wills 1997). The flotation of metallic values such as gold, silver, and copper in PCB CF can be compared to that of native occurrences of such metals, while for particles in end-use alloy forms, it is comparable to native alloys flotation (Nagaraj et al. 1992). From the conclusions in the work of Nagaraj et al. (1992), it is interesting for this context to observe that collector adsorption on alloy surface can be synergistic. Pure gold was observed to interact poorly with dicresyl monothiophosphate (DCMTP) except under a high oxidizing condition, which is not practical, whereas it adsorbs effectively on



113

gold–silver alloys over a wide range of conditions. This was found to be due to the strong silver–DCMTP interaction. It was shown by x-ray photoelectron and Fourier transform infrared spectroscopic techniques that the higher the percentage of silver alloyed with the gold, the more DCMTP was adsorbed. Hence, gold per se will not respond to DCMTP in practical situations, while Ag-bearing gold alloys will do. Therefore, it can be proposed that alloys in the PCB mix will respond to a collector as much as one of its constituent elements interacts with such a collector. This synergistic effect will enhance bulk metallic value collection from the fines. Although reports on flotation of metals have concentrated more on those that occur in native forms (which is relevant to real situations in classical minerals processing), many metals and alloys in the PCB fines that do not naturally occur in native forms can also be expected to respond to collectors. Auger electron spectroscopy (AES) and X-ray photoelectron spectroscopy (XPS) study of the chemisorbed xanthate monolayer on chalcocite and galena showed the same chemical environment for the metal atoms in the substrate and the monolayer xanthate (Buckley and Woods 1990, 1991; Shchukarev et al. 1994). This indicates interaction directly with the metal atoms in a substrate, implying that under adequate conditions the actual metal atom will interact with collectors and float. However, the potentials at which a collector compound will form on a pure metal and on its mineral compound are different (Woods 1996). Depending on the result of these schemes, selective followup floats can then be designed for the product fractions. Interestingly, qualitative exploratory investigations with microflotation cell have given indications that the schemes are feasible. A significantly high mass pull has been observed under the natural hydrophobic response. Gamma depression, as described previously, has also been observed to regulate this mass pull. Detailed quantitative investigations into the whole theme are in progress and reports will soon be available. 6. SURFACTANTS The schemes described will demand specific surface-conditioning reagents. Given the peculiar complexity of the PCB CF, the search will have to traverse the broad spectrum of surface-active agents– surfactants. Surfactants generally are organic compounds of heteropolar molecular structure. The nonpolar hydrocarbon chain group of the


114


molecule prefers to attach to air, while the polar functional groups prefer the aqueous phase. The surface activity derives from this property, as they can adsorb (accumulate) at air–water, air–mineral, and=or water–mineral interfaces. The specific application of a surfactant is determined by the properties of the polar functional group, which can be ionic–cationic or anionic, or nonionic. Examples of these compounds considered relevant in this context are shown in Table 4. Nonionic surfactants are nonelectrolytes, e.g. alcohols and paraffins, and they adsorb mainly at the air–water interface. The adsorption at the air—water interface lowers the surface tension of the solution, which makes thin films of the solution metastable and thus supports frothing, as applied in conventional minerals processing. It also creates wetting effects, lowering overall surface energy (surface tension). The reduction in the surface tension of the aqueous solution is toward that of the surfactant, and proportional to the concentration of the surfactant molecules adsorbed at the interface. This can be described by the Gibb’s adsorption equation in the form: X dc ¼ SS dT Ci dli ð1Þ i

Table 4. Common surfactant and applications Classification Ionic

Salts of organic acid

Example

Carboxylate salts Sodium alkly sulphonate Alkyl phosphate salt Amine salts Primary amine Sulphydryl Dithiocarbonates Trithiocarbonates Dithiophosphates Nonionic Alcohols Methyl IsoButyl Carbinol Pine oil Hydroxylated Polypropylene glycol polyglycol ethers methyl ether Tripolypropylene glycol methyl ether Nonylphenol polyglycol ether Alkoxy substituted Multi-ethoxylated butane parafins (R0 OH)xH (1,1,3 tri-ethoxy butane)

Major application Soap Wetting and detergency Wetting and detergency Collectors

Neutral frothers


115

where dc, SS, dT, Ci, and mi are the change in surface tension, surface entropy, change in temperature of the system, adsorption density of specie i, and the chemical potential of specie i, respectively, at constant pressure (Fuerstenau 1982). Defining i ¼ 1 for the solvent species, C1 ¼ 0 as the solvent does not adsorb in itself. Recalling that dmi ¼ RTd ln ai, at constant temperature (as in flotation process), it follows X X dc ¼ RT C1i d ln ai ﬃ RT C1i d ln Ci ð2Þ


i¼2

i¼2

where C1i refers to the surface excess (adsorption density) of specie i relative to the solvent at the interface and ai is the activity of i, which approximates the concentration of i, Ci, in very dilute solutions. The surface tension decreases toward a limiting value that is equal to that of the surfactant. It is noteworthy that Equation 2 is defined within the solubility range of the surfactant, with upper limit at its critical micelle concentration. The ionic surfactants are electrolytes and can adsorb at the air–water interface (also achieving some surface tension reduction), or at the particle–water interface electrostatically (physisorption) or chemically (chemisorption). Chemisorption, occurring at a given favorable pH and potential for the electrochemical reaction of the specific species, can be quite selective, as required of a good collector. For instance, xanthates will adsorb only on chalcocite at any pH greater than 7 in a pulp of pyrite and chalcocite. In the case of physisorption, as in heamatite flotation, the selectivity can be compromized, as two chemical species can have the same surface charge as the surfactant molecule, in a pulp so complex. Surface charge modification in such a system is also achieved by pH regulation, which will be more difficult with a multicomponent potential-pH system. For a system of more than 10 types of particles, as in this context, this may not be an easy solution. From the above brief, the selection of an appropriate surfactant for the present work will be easier. Gamma depression requires surface tension manipulation. Although surfactants generally lower surface tension, the choice should be narrowed down to the non-ionic ones. The activity (dissociation, adsorption) of ionic surfactants and other chemical depressants depends on many ions in the pulp, making it difficult to delineate or control the effect of such reagents addition in this pulp. A detailed investigation by Fraunholcz et al. (1997) on plastic wettability is quite useful in ruling out many other general classes of reagents for the depression. The work showed that plastic wettability appears to be

116


mainly governed by the critical surface tension concept as indicated with non-ionic frothers used in the work (Table 5). Cations (including Naþ, Ca2þ, Mg2þ, Fe2þ, Fe3þ, and Hþ) were said to have significant effect on the adsorption of non-ionic weak acids and the anionic

Table 5. Depressant investigated in the plastic wetting mechanism (Fraunholcz et al. 1997)


Class Nonionic surfactants (frothers)

Electrolytes

Inorganic depressants

Low-molecular-weight complex compounds

Macromolecular wetting agents

Name Diacetone alcohol Methyl isobutyl carbinol Tripolypropylene glycol methyl ether Iso-octanol Pine oil (a-Terpineol) Polyethylene glycol dodecyl Nonylphenol polyglycol ether. 10EO Na2CO3 NaCl NaNO3 MgCl2 CaCl2 FeCl2 FeCl3 HCl Sodium disulphite Sodium hexa meta phosphate Sodium silicate Malic acid Tartaric acid Citric acid 1,2,3-trihydroxy benzene 2,4,6-trihydroxy benzoic acid Arabicum gum, composition: m:n:p:q 1:1:3:3 Na-carboxymethyl cellulose Na-lignin sulfonate Potato starch Quebracho Tannic acid Polyvinyl alcohol Na-polyacrylic acid Polyacrylamide Polyethylene oxide



117

macromolecular depressants considered (Table 5). Low molecular weight complex compounds also do not adsorb sufficiently to effect depression, while inorganic reagents did not affect plastic flotation. Hence, such non-ionic surfactants as in Tables 4 and 5 will be candidates in PCB CF flotation, with further screening based on froth characteristics, effective dosages, availability, safety, costs, etc. On the other hand, the bulk metallic flotation scheme requires surfactants that distinctly chemisorb. The adsorption of xanthates (dithiocarbonates) on gold, platinum, silver, and copper was included in a detailed review of chemisorption of thiols on metals and metal sulphides by Woods (1996). Thus, xanthates will be good candidates for the scheme. To improve the bulk collection of the metallics, the selection can be narrowed to longer chain amyl xantathes and butyl xanthates. These are known to be bulk collectors in conventional minerals processing (Wills 1997). The basic mechanism is that hydrophobicity of the resultant surface is stronger with a longer hydrocarbon chain in the adsorbing collector; the longer the chain, the more positive the water—hydrocarbon interfacial free energy. With such long chain, little adsorption will suffice to impart enough hydrophobicity on metal particles with poorly adsorbing surface.

7. CONCLUSIONS The complex occurrences of eol PCB as a resource stock, which still continue to evolve more complexly in new generation devices, is a major constraint in its physical processing. The practice is getting close to three decades old and the challenges continue. The application of froth flotation to the fines generated in PCB comminution is considered to be promising and an overview has been presented on this hypothesis. It is recognized that many unknowns still exist, such as adequate regimes of kinetic parameters, pH modification of a mix so complex for selective chemisorption of theoretically appropriate collector, liberation, and realistic selectivity limits, reagents dosage regimes, and such. Investigations in this direction will be helpful for further clarifications. REFERENCES Alter, H., 2005, ‘‘The recovery of plastics from waste with reference to froth flotation.’’ Resources, Conservation and Recycling, 43, pp. 119–132.


118


Buchan, R. and Yara, B., 1995, ‘‘Recovering plastics for recycling by mineral processing techniques.’’ Journal of the Minerals, Metals and Materials Society, 43/4, pp. 52–55. Buckley, A. N. and Woods, R., 1990, ‘‘X-ray photoelectron spectroscopic and electrochemical studies of the interaction of xanthate with galena in relation to the mechanism proposed by Page and Hazell.’’ International Journal of Mineral Processing, 28(3–4), pp. 301–311. Buckley, A. N. and Woods, R., 1991, ‘‘Adsorption of ethyl xanthate on freshly exposed galena surfaces.’’ Colloids and Surfaces, 53, pp. 33–45. Castro, M. B., Remmerswaal, J. A. M., Brezet, J. C., van Schaik, A., and Reuter, M. A., 2005, ‘‘A simulation model of the comminution-liberation of recycling streams: Relationships between product design and the liberation of materials during recycling.’’ International Journal of Mineral Processing, 75, pp. 255–281. Coombs, C. F., 2001, Printed Circuits Handbook, 5th Ed., New York: Mc Graw Hill. D’Silva, K., 2004, ‘‘Brominated organic micropollutants—Igniting the flame retardants issue.’’ Critical Reviews in Environmental Science and Technology, 34, pp. 141–207. Dodbiba, G., Haruki, N., Shibayam, A., Miyazaki, T., and Fujita, T., 2002, ‘‘Combination of sink-float separation and flotation technique for purification of shredded PET-bottle from PE or PP flakes.’’ International Journal of Mineral Processing, 65, pp. 11–29. Elif, K. and Surendra, M. G., 2006, ‘‘Disassembly to order system under uncertainty.’’ Omega, 34=6, pp. 550–562. ESR International, 1994, Bidirectional dense medium barrel, US Patent 5,373,946 December 20, 1994. Feldmann, K. and Scheller, H., 1995, ‘‘The printed circuit board—A challenge for automated disassembly and for the design of recyclable interconnect devices.’’ CONCEPT—Conference on Clean Electronics Products and Technology, October 1995. IEE Conference Publication 415, pp. 186–190. Fraunholcz, N., 1997, ‘‘Plastics flotation.’’ Ph.D. Thesis, Delft University of Technology, The Netherlands. Fuersteanau, D. W., 1982, ‘‘Mineral-water interfaces and the electrical double layer,’’ In Principles of Flotation, (R. P. King, Ed.), Johannesburg, South Africa: SAIMM, pp. 17–30. Furter, L., 2004, ‘‘E-waster has dawned.’’ Re´source, May, pp. 8–11. Gold, S. and Dietz, K., 2003, ‘‘VOGT electronic FUBA makes its own rules,’’ Cicuitree, 16=2, pp. 28–34. Goosey, M. and Kellner, R., 2003, ‘‘Recycling technologies for the treatment of end of life printed circuit boards (PCBs).’’ Circuit World, 29, pp. 33–37. Goosey, M. and Kellner, R., 2002, ‘‘A scoping study: End-of –life printed circuit boards.’’ Department of Trade and Investment, Shipley Europe Ltd.



119

Hamos GmbH, Penzberg, Germany. Available at http://hamos.com/en/products/plants/scrap.htm (Assessed February 2008). Holm, E., 2001, ‘‘The printed circuit board primer.’’ Available at http://en. wikipedia.org/wiki/Printed_circuit_board#column-one (Assessed February 2008) and http://www.tomshardware.com/2001 /08/10/the_printed_circuit_ board_primer/ (Assessed February 2008). Iji, M. and Yokoyama, S., 1997, ‘‘Recycling of printed wiring boards with mounted electronic components.’’ Circuit World, 23=3, pp. 10–15. Jirang, C., 2005, ‘‘Mechanical recycling of consumer electronic scrap.’’ Licentiate thesis, Department of Chemical Engineering and Geosciences, Mineral Processing Division, Lulea University of Technology, Sweden. Kang H.-Y. and Schoenung, J. M., 2005, ‘‘Electronic waste recyling: A review of US infrastructure and technology options.’’ Resources, Conservation and Recycling, 45, pp. 368–400. Kellerwessel, H., 1993. ‘‘Concentration with air applications—Process, equipment, options, limits.’’ Aufbereitungs–Technik, 34=3, pp. 144–150. Kelly, G. E. and Spottiswood, D. J., 1989, Introduction to Mineral Processing, Perth, Australia: Wiley. Khalafalla, S. E., 1973, ‘‘Separating non-ferrous metals in incinerator residue using magnetic fluids.’’ Separation Science, 8, pp. 161–178. Khalafalla, S. E., 1976, ‘‘Magnetic separation of the second kind: magnetogravimetric, magnetohydrostatic, and magnetohydrodynamic separation.’’ IEEE Transactions on Magnetics, 12(5), pp. 455–462. Kogan, V., 2006, ‘‘Process for the recovery of precious metals scrap by means of hydrometallurgical technique.’’ Patent Application, WO 6006=013568 A3. Kravchenko, N. D., Dubinskii, Y. M., and Krichevskii, V. I., 1983, ‘‘Processing scrap of radio-electronic equipment.’’ The Soviet Journal of Non-Ferrous Metals (Tsvetnye Metally), 24, pp. 100–101. Laurie, J. F., 2005, ‘‘Poor nations are littered with old PCs.’’ New York Times, Oct. 24. Lee, C.-H., Chang-Tang, C., Kuo-Shuh, F., and Tien-Chin C., 2004, ‘‘An overview of recycling and treatment of scrap computers.’’ Journal of Hazardous Materials, B114, pp. 93–100. Lee, C.-H., Ssu-Li C., King-Min, W., and Lih-Chyi, W., 2000. ‘‘Management of scrap computer recycling in Taiwan.’’ Journal of Hazardous Materials, A73, pp. 209–220. Li, J., Puneet, S., Zong, G., and Hong-Chao, Z., 2004, ‘‘Printed circuit board recycling: a state of the art survey.’’ IEEE Transactions on Electronics Packaging Manufacturing, 27=1, pp. 33–42. Lungu, M., 2005. ‘‘Separation of small non-ferrous particles using an angular rotary drum eddy current separator with permanent magnets.’’ International Journal of Mineral Processing, 78, pp. 22–30.


120


Mackay, T., 2004, ‘‘Where do PCs go to die?’’ PC Format, Aug., pp. 16–20. Nagaraj, D. R., Brinen, J. S., Farinato, R. S., and Lee, J., 1992, ‘‘A study of interaction of Di-p-cresyl monothiophosphate with noble metals using electrochemical, wetting and spectroscopic methods.’’ Langmuir , 8(8), pp. 1943–1949. Nourreddine, M., Bjo¨rkman, B., Zhang, S. and Forssberg, E., 1998, ‘‘Thermodynamic conditions for the formation of dioxin during the recycling of non-ferrous metals from electric and elelctronic scrap’’ In EPC Congress, (B. Mishra, Ed.), Warrendale, PA: The Minerals, Metals and Materials Society. Ogunniyi, I. O. and Vermaak, M. K. G., 2007. ‘‘Improving printed circuit board physical processing—An overview.’’ Proceedings of EMC 2007, Dusseldorf. Pp. 1645–1656. Rosenqvist, T., 1988, Principles of Extractive Metallurgy, 2nd Ed., Auckland, New Zealand: McGraw Hill. Sander, S., Schubert, G., and Jackel, H.-G., 2004, ‘‘The fundamentals of the comminution of metals in shredders of the swing-hammer type.’’ International Journal of Mineral Processing, 74S, pp. 385–393. Shchukarev, A. V., Kravets, I. M., Buckley, A. N., and Woods, R., 1994. ‘‘Submonolayer adsorption of alkyl xanthates on galena.’’ International Journal of Mineral Processing, 41(1–2), pp. 99–114. Shent, H., Pugh, R. J., and Forssberg, E., 1999, ‘‘A review of plastics waste recycling and the flotation of plastics.’’ Resources, Conservation and Recycling, 25, pp. 85–109. Shuey, S. A., Vildal, E. E., and Taylor, P. R., 2006, ‘‘Pyrometallurgical processing of electronic waste.’’ Proceedings of the SME Annual Meeting, Preprint 06-037 pp. 1–8. March 27–29, St. Louis, MO. Stennett, A. D. and Whalley, D. C., 1999, ‘‘Novel techniques for electronic component removal.’’ Soldering and Surface Mount Technology, 11=2, pp. 7–11. Sum, E. Y. L., 1991, ‘‘The recovery of metals from electronic scrap.’’ Journal of the Minerals Metals and Materials Society, April, 43/4, pp. 53–60. Taylor, B., 2002, ‘‘Searching for solutions—Recycling electronic scrap.’’ Recycling Today, 40=6 56, pp. 58–60, 62. Vesilind, P. A. and Rimer, A. E., 1981. Unit Operations in Resource Recovery Engineering, Englewood Cliffs, NJ: Prentice Hall. Wikipedia, the free encyclopedia: Printed Circuit Board. Available at www. wikipedia.org Wills, B. A., 1997, Mineral Processing Technology, 6th Ed., London: Pergamon. Woods, R., 1996, ‘‘Chemisorption of thiols on metals and metals sulfides.’’ In Modern Aspects of Electrochemistry, # 29, (J. O. M. Bockris et al., Eds.), New York: Plenum, pp. 401–453. Yokoyama, S. and Iji, M., 1997, ‘‘Recycling of printed circuit board with mounted electronic part.’’ IEEE International Symposium on Electronics and the Environment, San Francisco, May, pp. 109–114.



121

Zhan, M. and Shenton, K. E., 1980, ‘‘Magnetic fluid bibliography.’’ IEEE Transactions on Magnetics, 16=2, pp. 387–409. Zhang, S. and Forssberg, E., 1997a, ‘‘Electronic scrap characterizaton for materials recycling.’’ Journal of Waste Management and Resource Recovery, 3, pp. 157–167. Zhang, S. and Forssberg, E., 1997b, ‘‘Mechanical separation-oriented characterization of electronic scrap.’’ Resource, Conservation and Recycling, 21, pp. 247–269. Zhang, S. and Forssberg, E., 1999, ‘‘Intelligent liberation and classification of electronic scrap.’’ Powder Technology, 105, pp. 295–301. Zhao, Y., Wen, X., Li, B., and Tao, D. 2004, ‘‘Recovery of copper from printed circuit boards.’’ Minerals and Metallurgical Processing, 21=2, pp. 99–102. Zisman, W. A., 1965, ‘‘Surface properties of plastics.’’ Record of Chemical Progress, 26, pp. 13–52.