Dining at the Periodic Table: Metals Concentrations as They Relate to

Feb 2, 2007 - Modelling global extraction, supply, price and depletion of the extractable geological resources with the LITHIUM model. Harald Ulrik Sv...
0 downloads 7 Views 389KB Size
Download PDF
Environ. Sci. Technol. 2007, 41, 1759-1765

Dining at the Periodic Table: Metals Concentrations as They Relate to Recycling J E R E M I A H J O H N S O N , * ,†,‡ E . M . H A R P E R , †,‡ R E I D L I F S E T , ‡ A N D T. E. GRAEDEL‡ Department of Chemical Engineering, and Center for Industrial Ecology, School of Forestry and Environmental Studies, Yale University, 205 Prospect St., Center for Industrial Ecology, School of Forestry and Environmental Studies, New Haven, Connecticut 06511

A correlation between the prices of a variety of substances and their dilutions in their initial matrices was shown in 1959 by T.K. Sherwood. The research presented here shows that the relationship holds for engineering metals today, which we termed the metals-specific Sherwood plot. The concentrations of metals in products (e.g., printed wiring boards and automobiles) and waste streams (e.g., municipal solid waste, and construction and demolition debris) were plotted with this correlation. In addition, for the products and waste streams that undergo disassembly at endof-life, the metals concentrations of the disassembled components were also plotted. It was found that most of the metals that are currently targeted for recycling have postdisassembly concentrations that lie above the metalsspecific Sherwood plot (i.e., have concentrations that are more enriched than minimum profitable ore grades). This suggests that material concentration plays a role in the viability of recycling at end-of-life. As products grow in complexity and the variety of materials used, analyses such as this one provide insight for policymakers and those interested in material sustainability into macro-level trends of material use and future recycling practices.

Introduction Increasing global population coupled with increases in consumption has led to dramatic increases in anthropogenic material flows. Mining rates for most metals are at historical maximums. While the increased availability of products often leads to increased standards of living, the related material flows are marked by such problems as high-energy consumption, natural resource depletion, and increased disposal. With the explosive global increase in metals use over time (see Figure 1) and a finite amount of virgin materials existing for human use, characterizing materials that could serve as resources for the future is a necessary and important task. In addition to the increasing magnitude of mass that is used, for many products there is an increasing variety of materials required that are contained in the components of products. This is certainly evident in the electronics industry, where we see product development and diversity of materials * Corresponding author phone: +1-203-506-9439; fax +1-203-4325556; e-mail:[email protected]. † Department of Chemical Engineering. ‡ Center for Industrial Ecology. 10.1021/es060736h CCC: $37.00 Published on Web 02/02/2007

 2007 American Chemical Society

FIGURE 1. Global production of 10 metals, 1900-2000; scaled 1900 rate ) 1.0, constructed with data from the United States Geological Survey, (1) and reprinted from Harper, et al., 2006 (2).

FIGURE 2. Use of elements in a circuit board (from T. McManus, Intel Corporation, private communication, 2006). advancing in tandem. One prime example is the evolution of the Intel circuit board, which in the 1980s contained 11 elements and in the 2000s will potentially balloon to over 60, as shown in Figure 2. (T. McManus, Intel Corporation, private communication, 2006) While reduction of toxicity and dematerialization may decrease the environmental impacts of products, once the products are produced, effective recycling can greatly improve life cycle environmental performance. Due to the increasing magnitude and diversity of material use, among other reasons, it is necessary to identify the drivers and barriers to recycling in the pursuit of sustainability. The present study focuses on metals concentrations in products and waste streams and recycling practices (including the role of disassembly). Our initial premise was to expand upon analyses previously completed by researchers. Concentration is certainly only one of the factors that influence the viability of recycling and we make no claim that concentration supersedes the role of minimum mass, dispersion, material purity, homogeneity, toxicity, or regulations. Instead, we aim to show that metals in products and waste streams that are recycled are predominately present in concentrations that exceed that of the traditional geological deposits that are currently utilized for the production of virgin resources. We believe that this analysis can inform policymakers and those studying the sustainability of materials. Specifically, as product and waste stream compositions evolve, this analysis can provide the basis for informed estimates regarding the implications of these changes. VOL. 41, NO. 5, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1759

FIGURE 3. (a) The relationship between materials dilution and price for virgin materials, as taken from Sherwood (5). (b) The relationship between materials dilution and price for virgin materials and industrial discards, from Allen and Behmanesh (4) (Reprinted with permission from The Greening of Industrial Ecosystems (1994) by the National Academy of Sciences, Courtesy of the National Academies Press, Washington, D.C.). An overarching goal of researchers in the field of industrial ecology (IE) is to understand and quantify the ways that people use materials. A major methodology guiding these efforts is material flow analysis (MFA). Most MFA work focuses on a nation or region and most studies to-date consist of static flow models for metals at the national and global levels for short time intervals (e.g., one year), with a smaller fraction of MFA studies incorporating dynamic models to analyze material flows over extended time intervals (3). Metals are often the focus of MFA and “design for environment” efforts due to the fact that metals are often repeatedly recyclable, are energy intensive to produce, often have associated toxicities (depending upon factors such as oxidation state), have substantial environmental consequences associated with their virgin extraction, can be recovered at or near highquality levels, and in many cases, have high unit value. Allen and Behmanesh produced a significant result in IE when they examined the potential of industrial discard streams for recycling (4). The selling prices of virgin materials are known to vary approximately proportionally with their degree of concentration in the matrix from which they are extracted, a result first published by Sherwood (5), as seen in Figure 3a. Sherwood’s plot considered a diverse suite of materials: vitamin B12, penicillin, copper, water, uranium, gold, and radium, with a regression line fit to the points. He asserted that separation constituted a significant portion of the costs of the pure substances, across a wide range of materials and separation mechanisms. King later published a modified and updated version of this plot and included several other compounds (6). Allen and Behmanesh compared this “Sherwood Plot” to the market prices of several metals plotted against the minimum concentration in 1760

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 5, 2007

discards that recyclers typically collect for recycling, as seen in Figure 3b. The line in the diagram represents the correlation between price and concentration as depicted in the original Sherwood plot and is not the regression line of the data points in this figure, which are concentrations of metals in recycled waste streams. They found that the loci for most of these metals lie above the Sherwood Plot; for these metals, the discard streams that are recycled are richer than traditional ore bodies. A major motivation of MFA is to characterize not just industrial discard streams but a variety of other discard streams for all global regions, providing a unique opportunity to expand upon Allen’s and Behmanesh’s initial foray (4). Because products are major constituents of discard streams they, too, may be analyzed in a similar way. We developed an updated Sherwood Plot, specific to the metals included in our analysis, which compare well to the most recent similar study conducted by Holland and Petersen in 1995 (7). The metals-specific Sherwood Plot is assumed to be applicable for all metals with a positive refined metal price. The class of substances to which the Sherwood Plot applies and especially factors that shape its applicability is, however, still a matter of investigation. As such, this paper is a preliminary reconnaissance. This plot allows rapid assessment of the dilution of the listed metals in various products as compared to their initial matrix, allowing for the determination that the metal is either enriched or diluted as compared to its initial source. Including price in this assessment, rather than simply looking at concentration, allows for greater understanding of the effect of price changes and, when assessed together, of how both the value of the metal and its concentration affect the desirability of recycling. We believe that the Sherwood plot is applicable for situations when there are massive amounts of unwanted material and/or when there are heterogeneous flows. In both cases, the key factor is the amount of target material relative to the total flow. Dahmus and Gutowski present an alternative approach using a Sherwood-type analysis, in which recycling of products is examined using the values of the materials and the mixture of them in various products (8, 9).

Methodology Metals Concentrations. Our first utilization of the Sherwood Plot (2) relied upon the previously published material dilution to price relationship (4). In the present work, the Sherwood Plot was recreated and made metals-specific, using current price and ore grade data for twenty of the most important engineering metals. Results are shown in Figure 4, with data sources detailed in Table 1. Obviously, a range of metal concentrations exists for ores currently mined, as does a range of prices at which the refined metals are sold. The ore grade ranges are illustrated for each metal on the plot, with price ranges given (where available) for the year 2004. To address the fact that several metals are co-mined with others, we attempted to isolate data on the range of operational ore grades in which the target metal was listed as the “primary” metal of the mine site and use only such sources. The Sherwood Plot (i.e., the regression fit to the points) was constructed using the minimum profitable ore grade for which each metal is mined and the average 2004 refined metal price. The recreated metals-specific Sherwood Plot was then used to analyze a variety of product composition data for four products: printed wiring boards, mobile phones, personal computers, and automobiles. A plot for each selected product was generated, showing some of the metals contained therein in tandem with the Sherwood Plot. Because no calculations were made to test the use of this correlation on materials not included in its formation, only the metals

product, and regulatory requirements shape the recycling rates. Once circuit boards are designated for recycling, however, we see similar metals targeted for reclamation, as shown in Figure 5a. Figure 5d and e show the effect of disassembly on metals concentrations and recycling. In Figure 5d, we see that many of the metals reside below the Sherwood line. When one separates the components of a car (with selected pieces detailed in Figure 5e), the metals concentrations rise, and we then see the link between concentration and recyclability. These figures suggest the effect of designing products for ease of disassembly.

FIGURE 4. The relationship between metals dilution in ore (the bar width) and 2004 price range (the bar height) for refined metal, with dilutions ranging from the lowest ore grade currently being mined (rightmost side of bars) to the highest (leftmost side of bars). The regression line is fit to the average 2004 refined metals prices and the lowest profitable ore grades. used to create the Sherwood Plot are included for each product. For example, silicon, an important component of computers, is not included in our analysis of the personal computer because this element was not used to develop the Sherwood Plot. However, nearly all of the metal content contained in the products examined is included in the twenty metals selected to produce Figure 4. A similar analysis was completed for several metals (copper, zinc, nickel, chromium, and silver) in discard flows, including construction and demolition debris (C&D), municipal solid waste (MSW), and waste from electronics and electrical equipment (WEEE). The role of disassembly was also examined by plotting the concentrations of components (post-disassembly) for the two examples that most readily see disassembly: automobiles and C&D. While “disassembly” is not common phraseology used in conjunction with C&D, it was used here because the result of separating its waste stream components is analogous to product disassembly.

Results Metals Concentrations. For several selected products, the Sherwood Plot is displayed in conjunction with product composition data. Figure 5a-e show the results for printed wiring boards, mobile phones, personal computers, automobiles, and automobile parts, with current recycling practices indicated in the figures. The data for the personal computer are representative of products from the mid-1990s (including a cathode ray tube display); computers of this vintage are currently entering end-of-life. Product composition data were obtained from a variety of reports and through personal communications, as indicated in the figure captions. Figure 5a-e demonstrate that, in a variety of products, the metals that are currently recycled to some degree predominately fall above the Sherwood line. An important caveat to this is shown for automobiles, in which case many of the metals that are recycled are in concentrations above the Sherwood line only after disassembly. As a general statement, this result indicates that when the “product ore” grade exceeds the minimum profitable grade of virgin ore, the “product ore” is commonly recovered and recycled. The exceptions are for low-value elements, such as tin in printed wiring boards. The recycling rates of circuit boards are dependent upon the product in which they are contained. Linked to that, issues such as mass of product, geographic dispersion of

This type of analysis was also completed for three discard flows: C&D, MSW, and WEEE, with results shown in Figure 5f-h. The metals concentrations in C&D, as shown in Figure 5f, were determined in a study in Zurich by Brunner and Stampfli (28). Details for Figure 5g-h are discussed in refs 29-33. The underlying analyses vary slightly from region to region due to data quality and availability. The analysis of copper and zinc in MSW applied varying metals concentrations to different components (e.g., organics and metal fractions) of the total discard flows. Silver, nickel, and chromium concentrations were applied to the bulk MSW flows, as a function of metal content in the bulk flow. The MSW data in this paper are derived from material that is collected as post-consumer waste, and it is important to note that, in some of the cases, materials recovery has taken place upstream (e.g., aluminum beverage containers). The analysis of WEEE began by applying a per capita WEEE generation rate to each country’s population, without the amount of WEEE recycled subtracted. For chromium, nickel, silver, and zinc, an average value was applied for each metal to the total WEEE flows. The analysis for copper in WEEE differed in that the total flows were subdivided into capital and consumer goods, because the copper concentration in capital flows is higher than that in consumer goods. Figure 5f demonstrates that (C&D) contains many metals at dilutions well below the Sherwood line, and recycling of these metals, therefore, are uneconomical. The main exceptions are two metalsscopper and ironswhich achieve a high concentration once they are removed from the waste stream, analogous to the disassembly of an automobile. Copper, which is present in pure form and easily recovered as wire and pipe, and iron, which is present in steel with sizable mass and a well-developed recycling infrastructure. The effect of the removal of this steel and copper is demonstrated using arrows in Figure 5f that indicate the increase in concentration. For MSW (Figure 5g), the dilutions are again well below the Sherwood line, and recovery and recycling are typically not attempted. WEEE (Figure 5h) is comprised of large white goods (e.g, refrigerators) and small electronics, such as the personal computers and mobile phones that have been discussed. All of these waste streams are heterogeneous in nature; if there are two very different WEEE streams, these concentrations may be applied to the Sherwood Plot and show very different results. A positive feature of this Sherwood-type analysis is that it rapidly isolates the materials that are richer than virgin ore at a given price. When analyzing the metal content of products and waste streams, it is important to consider the form and location of the metals. For effective recycling and disassembly, the metals must be accessible. Analysis, diagrams, and a brief discussion on the form and location of our target metals is available in Supporting Information, which we believe could serve as a foundation for broader discussions on the drivers to recycling. VOL. 41, NO. 5, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1761

TABLE 1. Ore Grade and Price Data Sources metal aluminum chromium

ore grade Sleppy, 2005 (10) Riekkola-Vanhanen, 1999 (12)

Raw Material Database (13) derived from global copper primary production germanium Scoyer et al., 2005 (14) gallium Greber, 2005 (15) gold Seymour and O'Farrelly, 2005 (16) iron Lepinski et al., 2005 (17) lead King et al., 2005 (18) manganese Matricardi and Downing (19) mercury DeVito and Brooks (20) molybdenum Stiefel, 2005 (21) nickel Raw Material Database, derived from global nickel primary production platinum Seymour and O'Farrelly, 2005 (16) palladium Seymour and O'Farrelly, 2005 (16) silver Hecla, 2005 (22); Glamis, 2005 (23) tin Graf, 2005 (24) titanium Infomine, 2006 (25) tungsten Penrice, 2005 (26) vanadium Bauer et al., 2005 (27) zinc Raw Material Database, derived from global zinc primary production

copper

Discussion By developing a current metals-specific Sherwood Plot and applying the concentrations of metals in products and discard streams, one can rapidly assess which metals are richer in these streams than they are in their virgin sources. The distance of a metal locus above the Sherwood line is related to recyclability, in that the higher a point is above the line, the more enriched the material, bringing with it increased potential for recycling. With few exceptions, it was demonstrated that metals in products and waste streams which are to some degree recycled are present in enriched concentrations (i.e., above the Sherwood line) in their post-disassembly form. Of the four products and three waste streams examined, there were 31 identified instances of metals recycling, of which 27 rested above the Sherwood plot post-disassembly. Two of the metals below the line were lead and mercury in personal computers, which are hazardous and regulated substances and, therefore, targeted for separation. The other instance in which a metal below the line was recycled was in the case of zinc in personal computers and automobiles, which may be explained by the ease of recycling galvanized steel. Although we see that most metals that are recycled lie above the Sherwood line, we do not see that all metals that lie above the Sherwood line are recycled. Among the four products and three waste streams, there are 39 instances where a metal lies above or on the Sherwood line, yet only 27 of these are, to some extent, recycled. This suggests that, while having an enriched concentration of a metal increases the likelihood of recycling, it is not sufficient to guarantee it. Concentration is by no means an isolated factor in determining whether a material is recycled. Other main factors that play into any recycling scheme are mass, dispersion of mass, contamination, homogeneity (43), and regulations. For a metal to be recycled, several hurdles must be overcome, including (1) A sufficient mass of the material must be available to economically justify collection and recycling processing infrastructure, and the concentration of the metal in the recycling stream must be high enough to make the separation processes economically viable, (2) The dispersion of the material (i.e., the distance between collection points) must not be prohibitive, (3) The impurity of the metal does not render recycling impractical. If the metal is bound or in alloy form, that will dictate the separation 1762

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 5, 2007

price USGS, 2005 (11) USGS, 2005 (price for ferrochromium, contained chromium) (11) USGS, 2005 (11) USGS, 2005 (price for electronic grade) (11) USGS, 2005 (11) USGS, 2005 (11) USGS, 2005 (2002 price for steel) (1) USGS, 2005 (11) USGS, 2005 (11) USGS, 2005 (11) USGS, 2005 (11) USGS, 2005 (11) USGS, 2005 (11) USGS, 2005 (11) USGS, 2005 (11) USGS, 2005 (Platts Metal Week composite) (11) USGS, 2005 (price for titanium sponge metal) (11) USGS, 2005 (price for concentrate, contained tungsten) (11) USGS, 2005 (price for vanadium pentoxide) (11) USGS, 2005 (11)

processes and costs associated with recovery, (4) The homogeneity of items to be recycled must be sufficient, (5) Depending on their structure and intent, regulations may mandate, encourage (e.g., bottle return laws), or discourage recycling. For those metals which reside above the Sherwood line but are not recycled, other factors such as contamination may have precluded their recycling. It is interesting to note that every material worth recovering in a mobile phone is located in the circuit board, and the recycling is driven by gold, silver, and palladium, as well as copper (personal communication: John Bullock, metals consultant). In Figure 5b, we see that these four metals are the ones farthest above the Sherwood line. A similar case is true for personal computers, in that some portion of every metal above the Sherwood Plot is typically recycled, with only two that are currently recycled falling below the plot (lead and zinc) (38). The “ease of disassembly” is an important factor when examining the concentration of metals in a product or waste stream; when disassembly readily occurs, it is essential to examine the concentrations of the components and not the whole product. An initial examination of the ease of disassembly of our products and waste streams showed that an advanced set of procedures exists for automotive disassembly, with some high-value parts dismantlers processing over 400 cars per day (40). In Figure 5d, most of the metals reside below the Sherwood line. However, many parts of the automobile are designed for rapid disassembly. By removing the lead acid battery, an action which often takes only several seconds, one would see the concentration of lead in the battery rise far above the Sherwood line (as shown for lead acid batteries in Figure 5e) and warrant recycling. Likewise, demolition crews are in the business of “disassembling” buildings, with the removal of high value items such as copper pipes making up an important revenue stream. On the other end of the spectrum, we see little disassembly for electronics, unless the goal is to remanufacture or reuse components. When the item, such as a mobile phone, has reached the true end of its useful life, if the phone reaches a recycler, it is shredded whole (possibly after the removal of the battery and liquid crystal display screen). This suggests that the disassembly is impractical or unnecessary.

FIGURE 5. Metals concentrations in relation to price for: (a) a printed wiring board (data averaged from three sources (34-36)); (b) a mobile phone (data from the Mobile Phone Partnership Initiative (37) and personal communication with John Bullock, metals consultant, 2006); (c) a personal computer (data from Pedersen, et al. (38)); (d) an average United States family automobile (data from Sullivan (39) and Staudinger, (40) with the exception of Hg (41)); (e) selected parts from a disassembled automobile (data from USGS, (42), Staudinger (40), and internal calculations); (f) C&D (data from a Zurich study by Brunner and Stampfli (28)); (g) MSW for selected regions (data from refs 29-33); (h) WEEE for selected regions (data from refs 29-33). CIS is Commonwealth of Independent States and LAC is Latin America and the Caribbean. VOL. 41, NO. 5, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1763

“Design for disassembly” can serve to facilitate reuse, remanufacture, and recycling. Whether this “design” is intentional is irrelevant, because the result is still the same: recyclers can disassemble such products and recycle or reuse their components. It is also true that the profitability of recycling one metal may drive the recycling of other metals that would not otherwise be targeted for recycling. The implication is that the disassembly (or the design for disassembly) of components not only affects the item in question, but also nontarget items. This study of recycling potentials of products and discard streams is properly regarded as no more than a reconnaissance: there remains much work to be done in this area. Recycling, and its associated economics, is a complex topic. All of the factors necessary for successful recycling demand a well-developed recycling infrastructure and processes, and the dispersion of mass necessarily entails collection systems planning. There are also other complicating factors, such as co-mined metals, where another metal is the primary metal of interest (for example, lead occurs as a co-product of zinc mining). Our plots for these materials do not capture the complexity of all of these attributes. If material scarcity becomes more pressing in the future, discard repositories such as landfills may increasingly be considered “mines”. An unpublished analysis showed that copper prices would have to increase by 40-fold for a landfill with 0.1% copper to be profitably mined for copper (personal communication: Gwendolyn McDay, Yale University, 2005). The material flows that add to discard repositories are the various discard streams. To that end, Figure 5h suggests that, on a purely mass dilution basis, mining WEEE discards may be a viable route for material recovery. Other discard streams (as shown in Figures 5f and 5g) appear less promising, but certainly cannot be disregarded out of hand, particularly if a material’s scarcity becomes very pressing. Should scarcity occur, we would likely see increased innovation, dematerialization, and substitution. All of these responses would have significant and interesting effects on this Sherwood analysis, and such scenario building is a potential extension of this work. We also recall the obvious: that if (or as) prices rise, metals previously below the Sherwood line will rise, eventually crossing above the intercept where recycling will be profitable. Massive mining and benefaction operations are designed to handle virgin sources, and capture economies of scale. Mining discard repositories would require innovative separation and recovery strategies. When comparing virgin and recycled resources, a complete analysis must take into account the economics of both processes. It may be the case that costs of mining and refining the anticipated lower-grade future virgin materials may eventually surpass the costs of mining discard streams and repositories. A Sherwood Plot analysis becomes increasingly relevant when considering the fact that products are becoming increasingly complex, their lifetimes are becoming shorter, and that global development is increasingly rapid (which means that global mine production, and all of its associated environmental burdens, is continually on the rise). The composition of various products and discard streams plotted in conjunction with the updated Sherwood Plot illustrate the potential for recycling and reuse in the end-of-life products that we typically call “waste”. We believe that there are four potential audiences for this research: recyclers, product designers, policymakers, and scientists who focus on material sustainability (e.g., industrial ecologists). Recycling companies are unlikely to utilize a tool such as this one unless it can identify profitable ventures that have not been identified. In its current state, a Sherwood analysis is too “blunt” of a tool to perform this task. With the pressing goals of cost reduction and improved functionality, 1764

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 5, 2007

it would be quite difficult for product designers to alter their designs with end-of-life characteristics taking precedence. The emergence of “design for environment” (44), product take-back, “design for disassembly” (45), and green engineering (46), however, may signal an budding paradigm shift that will encourage designers to use such tools. Nonetheless, the Sherwood Plot as outlined in this article could not be used in isolation by designers because metal concentration is not the only factor determining recyclability. Policymakers may note, for example, the success of a product easily disassembled (e.g., Figure 5e). By using policy tools, such as product take-back, they may encourage designers to work with product end-of-life in mind (47). Scientists who focus on material availability, resource use, and sustainability can use the Sherwood plot to evaluate design and recycling evolution over time and to assess macro-level trends in product design. In the case of both policymakers and sustainability scientists, when product design changes, the new metals concentrations may be plotted with the Sherwood Plot. Because it considers both economics and concentration, this may serve as a rough indicator of what metals in what products may enter future recycling streams. It may, in this way, serve to indicate future macro-level recycling trends. We think that this last use is likely to be the most compelling.

Acknowledgments We acknowledge support from the National Science Foundation (BES-9818788 and BES-0329470). We would also like to thank the United States Geological Survey Mineral Commodity Specialists for their effective and continued assistance and John Bullock for his consultation.

Supporting Information Available Additional details on the form and location of metals in the products analyzed in this paper are available. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) United States Geological Survey, Historical Statistics for Mineral and Material Commodities in the United States, Open-File Report 01-006, http://minerals.usgs.gov/minerals/pubs/of01006./, last accessed October 2005. (2) Harper, E. M.; Johnson, J.; Graedel, T. E. Making metals count: Applications of material flow analysis. Environ. Eng. Sci. 2006, 23, 493-506. (3) Fischer-Kowalski, M.; Hu¨ttler, W. Society’s metabolism: review of the intellectual history of materials flow analysis, Part II: 1970-1998. J. Ind. Ecol. 1998, 2, 107-135. (4) Allen, D.; Behmanesh, N. Wastes as raw materials, In The Greening of Industrial Ecosystems; Allenby, B., Richards, D., Eds.; National Academy Press: Washington, DC, 1994; pp 69-89. (5) Sherwood, T. K., “Mass Transfer Between Phases”, Phi Lambda Upsilon, Thirty-third annual Priestly Lectures, 1959. (6) King, C. J., Separation & Purification: Critical Needs and Opportunities; National Research Council (U.S.). Committee on Separation Science and Technology, National Academy Press: Washington, DC, 1987. (7) Holland, H. D.; Petersen, U. Living Dangerously: The Earth, its Resources, and the Environment; Princeton University Press: Princeton, NJ, 1995. (8) Gutowski, T. G.; Dahmus, J. B. Mixing Entropy and Product Recycling. Proceedings of the 2005 IEEE International Symposium on Electronics and the Environment; Institute of Electrical and Electronics Engineers: Piscataway, NJ, 2005, 72-76. (9) Dahmus, J. B.; Gutowski, T. G. Material Recycling at Product End-of-Life. Proceedings of the 2006 IEEE International Symposium on Electronics and the Environment; Institute of Electrical and Electronics Engineers: Piscataway, NJ, 2006, 206-211. (10) Sleppy, W. C. Aluminum Compounds, Introduction, In KirkOthmer Encyclopedia of Chemical Technology; John Wiley & Sons, Inc.: New York, 1992. (11) United States Geological Survey, Mineral Commodity Summaries, http://minerals.usgs.gov/minerals/pubs/commodity/, last accessed October 2005.

(12) Riekkola-Vanhanen, M., Finnish Expert Report on Best Available Techniques in Ferrochromium Production; The Finnish Environment Institute: Helsinki, 1999. (13) Raw Materials Data; Raw Materials Group: Stockholm, Sweden, 2005. (14) Scoyer, J.; Guislain, H.; Wolf, H. U. Germanium and Germanium Compounds, In Ullmann’s Encyclopedia of Industrial Chemistry; Wiley-VCH Verlag GmbH & Co.: Weinheim, Germany, 2002. (15) Greber, J. F. Gallium and Gallium Compounds, In Ullmann’s Encyclopedia of Industrial Chemistry; Wiley-VCH Verlag GmbH & Co.: Weinheim, Germany, 2002. (16) Seymour, R. J.; O’Farrelly, J. I. Platinum-Group Metals, In KirkOthmer Encyclopedia of Chemical Technology; John Wiley & Sons, Inc.: New York, 2001. (17) Lepinski, J. A.; Myers, J. C.; Geiger, G. H. Iron, In Kirk-Othmer Encyclopedia of Chemical Technology; John Wiley & Sons, Inc.: New York, 2005. (18) King, M.; Ramachandran, V.; Prengaman, R. D.; DeVito, S. C.; Breen, J. Lead and Lead Alloys, In Kirk-Othmer Encyclopedia of Chemical Technology; John Wiley & Sons, Inc.: New York, 2005. (19) Matricardi, L. R.; Downing, J. Manganese and Manganese Alloys, In Kirk-Othmer Encyclopedia of Chemical Technology; John Wiley & Sons, Inc.: New York, 2005. (20) DeVito, S. C.; Brooks, W. E. Mercury, In Kirk-Othmer Encyclopedia of Chemical Technology; John Wiley & Sons, Inc.: New York, 2005. (21) Stiefel, E. I. Molybdenum and Molybdenum Alloys, In KirkOthmer Encyclopedia of Chemical Technology; John Wiley & Sons, Inc.: New York, 2001. (22) Hecla, http://www.hecla-mining.com/mainProperties.html, last accessed May 2005. (23) Glamis, http://www.glamis.com/corporate/ore.html, last accessed May 2005. (24) Graf, G. G. Tin, Tin Alloys, and Tin Compounds, In Ullmann’s Encyclopedia of Industrial Chemistry; Wiley-VCH Verlag GmbH & Co.: Weinheim, 2002. (25) Infomine, Randsburg Int’l cuts 589 m of Fe, Ti and V at Titan, http://www.infomine.com/index/pr/PA316365.PDF, last accessed January 6, 2006. (26) Penrice, T. W. Tungsten and Tungsten Alloys, In Kirk-Othmer Encyclopedia of Chemical Technology; John Wiley & Sons, Inc.: New York, 2001. (27) Bauer, G.; Gu ¨ ther, V.; Hess, H.; Otto, A.; Roidl, O.; Roller, H.; Sattelberger, S. Vanadium and Vanadium Compounds, In Ullman’s Encyclopedia of Industrial Chemistry; Wiley-VCH Verlag GmbH & Co.; John Wiley & Sons, Inc.: New York, 2002. (28) Brunner, P. H.; Stampfli, D., The material balances of a construction waste sorting plant. Waste Manage. Res. 1993, 11, 27-48. (29) Harper, E. M.; Bertram, M.; Graedel, T. E., The contemporary Latin America and Caribbean zinc cycle: One year stocks and flows. Resour. Conserv. Recycle 2006, 47, 82-100. (30) Bertram, M.; Graedel, T. E.; Rechberger, H.; Spatari, S., The contemporary European copper cycle: waste management subsystem. Ecol. Econ. 2002, 42, 43-57. (31) Johnson, J.; Jirikowic, J.; Bertram, M.; Beers, D. v.; Gordon, R. B.; Henderson, K.; Klee, R. J.; Lanzano, T.; Lifset, R.; Oetjen, L.; Graedel, T. E., Contemporary anthropogenic silver cycle: A multilevel analysis. Environ. Sci. Technol. 2005, 39, 4655-4665.

(32) Johnson, J.; Schewel, L.; Graedel, T. E., The contemporary anthropogenic chromium cycle. Environ. Sci. Technol. 2006, 40, 7060-7069. (33) Reck, B.; Mueller, D. B.; Rostkowski, K.; Graedel, T. E. Characterization of the contemporary anthropogenic nickel cycle 2006, in preparation. (34) Salazar, C., Semi-conductors from the Philippines, http:// www.iisd.org/susprod/semiconduct.pdf, last accessed November 2, 2005. (35) Goosey, M.; Kellner, R., A Scoping Study End-of-Life Printed Circuit Boards, http://www.intellectuk.org/sectors/components_ and_manufacturing/environmental/end_of_life_pcb_study.pdf, last accessed November 2, 2005. (36) Veit, H.; Pereira, C.; Bernardes, A., Using Mechanical Processing in Recycling Printed Wiring Boards. JOM 2002, 54, 45-47. (37) Mobile Phone Partnership Initiative, Environmentally Sound Management of End-of-Life Mobile Phones, The Basel Convention on the Control of Transboundary Movements of Hazardous Wastes and their Disposal 2004. (38) Pedersen, S.; Wilson, C.; Pitts, G.; Stotesbery, B., Electronics Industry Environmental Roadmap; MCC TECHNICAL REPORT: MCC-ECESM-001-96; Microelectronics and Computer Technology Corp.: Austin, TX, 1996. (39) Sullivan, J. L.; Williams, R. L.; Yester, S.; Cobas-Flores, E.; Chubbs, S. T.; Hentges, S. G.; Pomper, S. D., Life Cycle Inventory of a Generic US Family SedansOverview of Results, USCAR AMP Project, paper 982160; Society of Automobile Engineers: Warrendale, PA, 1998. (40) Staudinger, J.; Keoleian, G. A., Management of End-of Life Vehicles (ELVs) in the US, report no. CSS01-01; University of Michigan: Ann Arbor, MI, 2001. (41) Griffith, C.; Gearhart, J.; Posselt, H.; McPherson, A.; Gingras, S.; Davis, G.; Dhingra, R.; Kincaid, L., Toxics in Vehicles: Mercury; The Ecology Center, Great Lakes Unites, and the University of Tennessee Center for Clean Products and Clean Technologies, 2001. (42) Hilliard, H., Platinum Recycling in the United States in 1998, USGS Survey Circular 1196-B; United States Geological Survey: Reston, VA, 1998. (43) Quimby, T., Recycling: the Alternative to Disposal; Johns Hopkins University Press for Resources for the Future: Baltimore, MD, 1975. (44) Billatos, S. B.; Basaly, N. A., Green Technology and Design for the Environment; Taylor & Francis: Washington, DC, 1997. (45) Graedel, T. E.; Allenby, B. R., Design for Environment; Prentice Hall: Upper Saddle River, NJ, 1996. (46) Anastas, P.; Zimmerman, J., Design through the 12 principles of green engineering. Environ. Sci. Technol. 2003, 37, 94A-101A. (47) Lindhqvist, T.; Lifset., R., Getting the goal right: EPR and DfE. J. Ind. Ecol. 1998, 2, 6-8.

Received for review March 28, 2006. Revised manuscript received December 12, 2006. Accepted December 18, 2006. ES060736H

VOL. 41, NO. 5, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1765