noble-metals recovery from printed circuit boards: a
NOBLE-METALS RECOVERY FROM PRINTED CIRCUIT BOARDS: A MULTIDISCIPLINARY APPROACH TOWARDS SUSTAINABILITY A. RIGOLDIa, E. F. TROGUa, G. C. MARCHESELLIb, N. PICONEc, P. DEPLANOd, M. COLLEDANIe, AND A. SERPEa,f* a
Dipartimento di Scienze Chimiche e Geologiche, University of Cagliari, SS 554 Bivio per Sestu, 09042 Monserrato (CA), Italy b Fabbrica Italiana Leghe Metalliche Sinterizzate (F.I.L.M.S.) SpA, Via Megolo 49, 28877 Anzola D'Ossola (VB), Italy c ITIA-CNR, Istituto Tecnologie Industriali ed Automazione, Via Corti 12, 20100, Milano, Italy d Dipartimento di Fisica, University of Cagliari, SS 554 Bivio per Sestu, 09042 Monserrato (CA), Italy e Politecnico di Milano, Dipartimento di Ingegneria Meccanica, Via la Masa 1, 20157, Milano, Italy f Dipartimento di Ingegneria Civile, Ambientale ed Architettura (DICAAR), Via Marengo 2, 09123 Cagliari, Italy
SUMMARY: A robust database of the noble metals (NM) content of different classes of Printed Circuit Boards (PCB) was build through an accurate characterization of samples of different origin, underwent mechanical comminution and representative sampling, by ICP-AES quantitative chemical analysis, to work as reference for researcher and companies interested in waste valorisation. The results identified RAM and mobile phone’s PCB as the “richest” classes of PCB, while TV PCB as the “poorest” in term of NM content. On these basis, a sustainable three-step NM recovery method, previously set-up on a finely shredded WEEE sample deprived by non-metallic and ferrous materials, was applied on a coarse sample of shredded RAM boards provided by companies, as a case study. Preliminary results highlighted that good NM recovery yields (from 65% to quantitative) and limited by-products formation could be achieved, despite a huge amount of composite materials was present in the mixture. Nevertheless, 10times increased leaching times found for copper dissolution suggested that improved mechanical pretreatments might increase leaching effectiveness.
1. INTRODUCTION Noble metals (NM) represent relevant high-value components of WEEE (Waste Electric and Electronic Equipments) and their recovery makes these scraps an intriguing source of valuable raw materials. Among the valuable materials contained in WEEE, Printed Circuit Boards (PCB) represent a real “urban mine” of key metals, containing amount of precious and “critical” metals Proceedings Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium/ 2 - 6 October 2017 S. Margherita di Pula, Cagliari, Italy / © 2017 by CISA Publisher, Italy
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
often larger than in their ores as reported in several papers (Tuncuk et al, 2012; Yamane et al., 2011; Park and Fray, 2009; Yang et al., 2009, Guo et al., 2009, Goosey and Kellner, 2002; Sum, 2005). Nevertheless, a systematic quantitative characterization of the valued metal content in different kind of PCB is still lacking making it difficult to assess the profitability of their end-of-life valorization. Despite the growing interest in waste valorization for both economical and environmental purposes, also due to the recent world-wide regulations in terms of waste management and prevention of resource depletion, NM are still scarcely recovered and the recovery is performed only in few countries where companies, mostly already operating in the field of NM production and/or treatment by means of conventional methods, adapted their industrial process to the recovery pourposes. The most diffused industrial practices for NM recovery from PCB are based on pyrolysis and electrolysis often involving very harmful and/or energy-demanding processes (Cui and Zhang, 2008). Hydrometallurgical methods, deriving from the mineral industry are also industrially applied, in particular for NM recycling form selected WEEE like PCB, but their application is limited by the use of harmuful reagents (i.e. cynidation and aqua-regia leaching)(Tuncuk, 2012) and by the heterogeneous and composite composition of the scraps which makes the process less effective and selective. The high costs and, often, harmfulness of the recovering processes, heavely limit the entry into the market of new actors interested in NM recovery. Nevertheless more sustainable NM recovery from complex matrix as PCB is a very challenging issue not satisfactorily solved yet. In this context, new sustainable hydrometallurgical methods recently set-up by us seem to give a promising answer with the view to design new industrial plants able to combine effectiveness and low environmental impact. In particular, we found high effectiveness and selectivity, in high purity NM recovery by using safe, selective and recyclable reagents in very mild conditions from a mixed metallic fraction of WEEE obtained by means of shredding and electrostatic and magnetic separation techniques (Serpe et al., 2015; Serpe et al. 2014). Nevertheless the very inhomogeneous composition of WEEE and the inefficiency of commercial mechanical pretreatments plants, in terms of loss of valuable materials during the shredding and separation phases and their cost, still limit the industrial application of sustainable NM recovery methods. It is worth noting that the number of companies able to produce on their own a raw shredded material is greatly increasing, while just few of them are currently able to produce a good quality semi-finished metallic material obtained as the output of a selective mechanical separative process. On these basis, in order to promote the NM recovery from PCB and the technology transfer of sustainable hydrometallurgical processes, here we report: i) the accurate gold, silver, palladium and copper characterization by means of quantitative chemical analysis, by a selection of comminuted PCB samples underwent combined mechanical (shredding) and chemical (leaching) pretreatments, in order to build a robust database useful as reference for researcher and companies interested in waste valorisation; and, as a case study, ii) the preliminary results on the application of our sustainable NM recovery process on a shredded RAM sample provided by companies, highly heterogeneous in terms of size distribution and composition, in order to test the selectivity and effectiveness of the method on a very coars material and when the vitreous-plastic support and the ferrous materials are still present in the mixture. 2. MATERIALS AND METHODS Reagents and solvents for chemical digestion, characterization and leaching of the samples, were purchased from Sigma-Aldrich (except citric acid anhydrous FU-E 330 from ACEF, Italy) and used without further purification. PCB of different origin were supplied by companies (see aknowledgments).
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
2.1 WEEE comminution and characterization 2.1.1 Comminution The size-reduction of PCB was performed in dry condition in two different stages after the removal of some minor components (i.e. motor and rotating supports, fasteners and ferrous parts), without loss of NM, to avoid machine jamming and damages. A preliminary comminution was performed with a single shaft shredder. Given the heterogeneous nature of samples, they were treated separately and internal cleaning of the machine was performed after each treatment, in order to avoid contamination and loss of precious materials. The output product of this preliminary size-reduction (10 mm) was thus subjected to a typical shredding process with cutting mill (1500 rpm; meshed grid: 2 mm), by using cutting tools of hardmetal (WC-based materials). 2.1.2 Sampling Given the heterogeneous nature of the material, all fractions of the same type of sample were gathered into a single container and underwent homogenization. Sampling was made by coring in different regions of the sample, in order to obtain representative aliquots for the subsequent treatments. In particular, for each sample: A. n. 5 1g aliquots underwent microwave mineralization; B. n. 1 20g aliquot underwent acid digestion under room conditions. 2.1.3 Sample digestion The aliquots referred to in point A, underwent microwave mineralization by HNO3 (65%, 2 mL), HCl (37%, 6 mL), H2O2 (30%, 0.5 mL) mixture and treatment in a Milestone Ethos 1 Microwave digester, applying a microwave program consisting of two steps lasting 10 and 20 minutes, respectively, at a temperature of 220 °C and microwave power up to 1000 W. Differently, the bulky corings referred to in point B, were subjected to prolonged digestion (24h) in a open flask at room temperature, preliminarily by HNO3 65% and subsequently by aquaregia.
2.1.4 ICP-AES characterization The samples were prepared for the analysis by diluting the digested solutions with 1% HNO3 blank and then the metals were determined using a Varian Liberty 200 Inductively Coupled Plasma-Atomic Emission Spectrometer (ICP-AES), with respect to 5-point calibration plots in the 1–50 ppm range for copper and in the 1–10 ppm range for the other metals (correlation coefficient for single element calibration line > 0.9999). 2.2 NM leaching on RAM #R#G#M1 sample 2.2.1 RAM milling 200 g of mix RAM boards were milled for 24h in a stainless steel jar by means of hardmetal balls (Ø 6 mm, 1.4Kg) in a planetary apparatus (4-stages Retsch mill, 300rpm) in the presence of Carbsyn (250mL). At the end of the process, Carbsyn was recovered by distillation and the milled sample dried. 1
Selected sample of the project: #Recovery #Green #Metal; supported by companies and interested people through the crowdfunding campaign withyouwedo (see acknoledgments).
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
2.2.2 Characterization of the sample N. 4 aliquots of about 1.5g each of RAM #R#G#M sample were digested under microwaves,and characterized by ICP-AES as described in paragraphs 2.1.3 and 2.1.4. 2.2.3 Leaching and recovery processes Selective leaching and NM recovery procedure was applied as fully described in the litterature (Serpe et al., 2015) on 10g of the raw sample described above. Electrowinning experiments for metal recovery were performed by an apparatus constituted by two metallic electrodes (Pt-wire for the anode and Cu-wire for the cathode) connected, to their perspective ends, with an external electron supply (Thurlby Thandar Instruments, mod. PL 310) which drives the process.
3. RESULTS AND DISCUSSION 3.1 Samples’ classification, comminution and characterization 3.1.1 Physical processes PCB of different origin were supplied by Gold Fixing and Ri.Tech companies in the number, weight and typologies summarized in Table 1, with the view to build a knowledge database useful to select the most promising PCB typologies for a sustainable NM recovery. Table 1. List of PCB samples underwent comminution.
a
b
c
Network Integrated Controllers; Random Access Memories; Shredded with connectors
In order to obtain the most accurate results, all available PCB of each different typology were gathered and submitted to comminution, preventing material loss as much as possible (maximum materials loss accounted for 2.9%). Separative procedures were avoided in order to
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
limit the loss of NM which is tipically observed during eddy current and magnetic separations1. 3.1.2 Sampling and characterization In order to obtain representative and accurate results, the whole amount of sample for each PCB typology was collected in a flat bottom box, homogenized and leveled as much as possible. Then it was sampled by quarting the material and by coring in order to overcome material layering, obtaining 5 different aliquots of about 1g each. The different aliquots were then digested under microwaves by a mixture of HNO3, HCl and H2O2 as detailed in the Materials and Methods section. At the same time, larger amounts (20g) of the same samples were massively sampled and leached in two steps by means of HNO3 65% solution, followed by an aqua-reagia treatment. Finally all the digested fractions were properly diluted by a 1% HNO3 blank and measured for Au, Cu, Ag and Pd over five-points calibration plots.The results obtained for the two kind of sampling and treatments agreed. Table 2 summarizes the metal content percentages, calculated as weighted average of the measured values by the different analysed aliquots. Table 2. NM content of different typologies of PCB determined by ICP-AES measurements. Sample DVD/CD players PCBs NICs RAM white fingers RAM gold fingers Hard drive PCBs Mother boards TV PCBs Mobile phone PCBs
Pd
Dipartimento di Scienze Chimiche e Geologiche, University of Cagliari, SS 554 Bivio per Sestu, 09042 Monserrato (CA), Italy b Fabbrica Italiana Leghe Metalliche Sinterizzate (F.I.L.M.S.) SpA, Via Megolo 49, 28877 Anzola D'Ossola (VB), Italy c ITIA-CNR, Istituto Tecnologie Industriali ed Automazione, Via Corti 12, 20100, Milano, Italy d Dipartimento di Fisica, University of Cagliari, SS 554 Bivio per Sestu, 09042 Monserrato (CA), Italy e Politecnico di Milano, Dipartimento di Ingegneria Meccanica, Via la Masa 1, 20157, Milano, Italy f Dipartimento di Ingegneria Civile, Ambientale ed Architettura (DICAAR), Via Marengo 2, 09123 Cagliari, Italy
SUMMARY: A robust database of the noble metals (NM) content of different classes of Printed Circuit Boards (PCB) was build through an accurate characterization of samples of different origin, underwent mechanical comminution and representative sampling, by ICP-AES quantitative chemical analysis, to work as reference for researcher and companies interested in waste valorisation. The results identified RAM and mobile phone’s PCB as the “richest” classes of PCB, while TV PCB as the “poorest” in term of NM content. On these basis, a sustainable three-step NM recovery method, previously set-up on a finely shredded WEEE sample deprived by non-metallic and ferrous materials, was applied on a coarse sample of shredded RAM boards provided by companies, as a case study. Preliminary results highlighted that good NM recovery yields (from 65% to quantitative) and limited by-products formation could be achieved, despite a huge amount of composite materials was present in the mixture. Nevertheless, 10times increased leaching times found for copper dissolution suggested that improved mechanical pretreatments might increase leaching effectiveness.
1. INTRODUCTION Noble metals (NM) represent relevant high-value components of WEEE (Waste Electric and Electronic Equipments) and their recovery makes these scraps an intriguing source of valuable raw materials. Among the valuable materials contained in WEEE, Printed Circuit Boards (PCB) represent a real “urban mine” of key metals, containing amount of precious and “critical” metals Proceedings Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium/ 2 - 6 October 2017 S. Margherita di Pula, Cagliari, Italy / © 2017 by CISA Publisher, Italy
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
often larger than in their ores as reported in several papers (Tuncuk et al, 2012; Yamane et al., 2011; Park and Fray, 2009; Yang et al., 2009, Guo et al., 2009, Goosey and Kellner, 2002; Sum, 2005). Nevertheless, a systematic quantitative characterization of the valued metal content in different kind of PCB is still lacking making it difficult to assess the profitability of their end-of-life valorization. Despite the growing interest in waste valorization for both economical and environmental purposes, also due to the recent world-wide regulations in terms of waste management and prevention of resource depletion, NM are still scarcely recovered and the recovery is performed only in few countries where companies, mostly already operating in the field of NM production and/or treatment by means of conventional methods, adapted their industrial process to the recovery pourposes. The most diffused industrial practices for NM recovery from PCB are based on pyrolysis and electrolysis often involving very harmful and/or energy-demanding processes (Cui and Zhang, 2008). Hydrometallurgical methods, deriving from the mineral industry are also industrially applied, in particular for NM recycling form selected WEEE like PCB, but their application is limited by the use of harmuful reagents (i.e. cynidation and aqua-regia leaching)(Tuncuk, 2012) and by the heterogeneous and composite composition of the scraps which makes the process less effective and selective. The high costs and, often, harmfulness of the recovering processes, heavely limit the entry into the market of new actors interested in NM recovery. Nevertheless more sustainable NM recovery from complex matrix as PCB is a very challenging issue not satisfactorily solved yet. In this context, new sustainable hydrometallurgical methods recently set-up by us seem to give a promising answer with the view to design new industrial plants able to combine effectiveness and low environmental impact. In particular, we found high effectiveness and selectivity, in high purity NM recovery by using safe, selective and recyclable reagents in very mild conditions from a mixed metallic fraction of WEEE obtained by means of shredding and electrostatic and magnetic separation techniques (Serpe et al., 2015; Serpe et al. 2014). Nevertheless the very inhomogeneous composition of WEEE and the inefficiency of commercial mechanical pretreatments plants, in terms of loss of valuable materials during the shredding and separation phases and their cost, still limit the industrial application of sustainable NM recovery methods. It is worth noting that the number of companies able to produce on their own a raw shredded material is greatly increasing, while just few of them are currently able to produce a good quality semi-finished metallic material obtained as the output of a selective mechanical separative process. On these basis, in order to promote the NM recovery from PCB and the technology transfer of sustainable hydrometallurgical processes, here we report: i) the accurate gold, silver, palladium and copper characterization by means of quantitative chemical analysis, by a selection of comminuted PCB samples underwent combined mechanical (shredding) and chemical (leaching) pretreatments, in order to build a robust database useful as reference for researcher and companies interested in waste valorisation; and, as a case study, ii) the preliminary results on the application of our sustainable NM recovery process on a shredded RAM sample provided by companies, highly heterogeneous in terms of size distribution and composition, in order to test the selectivity and effectiveness of the method on a very coars material and when the vitreous-plastic support and the ferrous materials are still present in the mixture. 2. MATERIALS AND METHODS Reagents and solvents for chemical digestion, characterization and leaching of the samples, were purchased from Sigma-Aldrich (except citric acid anhydrous FU-E 330 from ACEF, Italy) and used without further purification. PCB of different origin were supplied by companies (see aknowledgments).
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
2.1 WEEE comminution and characterization 2.1.1 Comminution The size-reduction of PCB was performed in dry condition in two different stages after the removal of some minor components (i.e. motor and rotating supports, fasteners and ferrous parts), without loss of NM, to avoid machine jamming and damages. A preliminary comminution was performed with a single shaft shredder. Given the heterogeneous nature of samples, they were treated separately and internal cleaning of the machine was performed after each treatment, in order to avoid contamination and loss of precious materials. The output product of this preliminary size-reduction (10 mm) was thus subjected to a typical shredding process with cutting mill (1500 rpm; meshed grid: 2 mm), by using cutting tools of hardmetal (WC-based materials). 2.1.2 Sampling Given the heterogeneous nature of the material, all fractions of the same type of sample were gathered into a single container and underwent homogenization. Sampling was made by coring in different regions of the sample, in order to obtain representative aliquots for the subsequent treatments. In particular, for each sample: A. n. 5 1g aliquots underwent microwave mineralization; B. n. 1 20g aliquot underwent acid digestion under room conditions. 2.1.3 Sample digestion The aliquots referred to in point A, underwent microwave mineralization by HNO3 (65%, 2 mL), HCl (37%, 6 mL), H2O2 (30%, 0.5 mL) mixture and treatment in a Milestone Ethos 1 Microwave digester, applying a microwave program consisting of two steps lasting 10 and 20 minutes, respectively, at a temperature of 220 °C and microwave power up to 1000 W. Differently, the bulky corings referred to in point B, were subjected to prolonged digestion (24h) in a open flask at room temperature, preliminarily by HNO3 65% and subsequently by aquaregia.
2.1.4 ICP-AES characterization The samples were prepared for the analysis by diluting the digested solutions with 1% HNO3 blank and then the metals were determined using a Varian Liberty 200 Inductively Coupled Plasma-Atomic Emission Spectrometer (ICP-AES), with respect to 5-point calibration plots in the 1–50 ppm range for copper and in the 1–10 ppm range for the other metals (correlation coefficient for single element calibration line > 0.9999). 2.2 NM leaching on RAM #R#G#M1 sample 2.2.1 RAM milling 200 g of mix RAM boards were milled for 24h in a stainless steel jar by means of hardmetal balls (Ø 6 mm, 1.4Kg) in a planetary apparatus (4-stages Retsch mill, 300rpm) in the presence of Carbsyn (250mL). At the end of the process, Carbsyn was recovered by distillation and the milled sample dried. 1
Selected sample of the project: #Recovery #Green #Metal; supported by companies and interested people through the crowdfunding campaign withyouwedo (see acknoledgments).
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
2.2.2 Characterization of the sample N. 4 aliquots of about 1.5g each of RAM #R#G#M sample were digested under microwaves,and characterized by ICP-AES as described in paragraphs 2.1.3 and 2.1.4. 2.2.3 Leaching and recovery processes Selective leaching and NM recovery procedure was applied as fully described in the litterature (Serpe et al., 2015) on 10g of the raw sample described above. Electrowinning experiments for metal recovery were performed by an apparatus constituted by two metallic electrodes (Pt-wire for the anode and Cu-wire for the cathode) connected, to their perspective ends, with an external electron supply (Thurlby Thandar Instruments, mod. PL 310) which drives the process.
3. RESULTS AND DISCUSSION 3.1 Samples’ classification, comminution and characterization 3.1.1 Physical processes PCB of different origin were supplied by Gold Fixing and Ri.Tech companies in the number, weight and typologies summarized in Table 1, with the view to build a knowledge database useful to select the most promising PCB typologies for a sustainable NM recovery. Table 1. List of PCB samples underwent comminution.
a
b
c
Network Integrated Controllers; Random Access Memories; Shredded with connectors
In order to obtain the most accurate results, all available PCB of each different typology were gathered and submitted to comminution, preventing material loss as much as possible (maximum materials loss accounted for 2.9%). Separative procedures were avoided in order to
Sardinia 2017 / Sixteenth International Waste Management and Landfill Symposium / 2 - 6 October 2017
limit the loss of NM which is tipically observed during eddy current and magnetic separations1. 3.1.2 Sampling and characterization In order to obtain representative and accurate results, the whole amount of sample for each PCB typology was collected in a flat bottom box, homogenized and leveled as much as possible. Then it was sampled by quarting the material and by coring in order to overcome material layering, obtaining 5 different aliquots of about 1g each. The different aliquots were then digested under microwaves by a mixture of HNO3, HCl and H2O2 as detailed in the Materials and Methods section. At the same time, larger amounts (20g) of the same samples were massively sampled and leached in two steps by means of HNO3 65% solution, followed by an aqua-reagia treatment. Finally all the digested fractions were properly diluted by a 1% HNO3 blank and measured for Au, Cu, Ag and Pd over five-points calibration plots.The results obtained for the two kind of sampling and treatments agreed. Table 2 summarizes the metal content percentages, calculated as weighted average of the measured values by the different analysed aliquots. Table 2. NM content of different typologies of PCB determined by ICP-AES measurements. Sample DVD/CD players PCBs NICs RAM white fingers RAM gold fingers Hard drive PCBs Mother boards TV PCBs Mobile phone PCBs
Pd
