Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2025 Sep 15;59(37):19755–19767. doi: 10.1021/acs.est.5c04952

Advancing the Economic and Environmental Sustainability of Rare Earth Element Recovery from Phosphogypsum

Adam Smerigan a,b, Rui Shi a,c,*
PMCID: PMC12461911  PMID: 40948472

Abstract

Transitioning to green energy requires more sustainable rare earth element (REE) production. The current REE supply relies on energy- and chemical-intensive mining, prompting interest in alternative sources like phosphogypsum (PG) waste. However, using conventional solvent extraction to recover REEs in PG is inefficient and environmentally burdensome. This study proposes a treatment train for REE recovery from PG, featuring a bioinspired adsorptive separation, and evaluates its environmental and economic performance using a probabilistic sustainability framework that integrates life cycle assessment (LCA) and techno-economic analysis (TEA). Results show the system achieves an internal rate of return (IRR) above 15% in 87% of simulations, suggesting strong profitability potential. Environmentally, it outperforms conventional REE mining and PG treatment in ecosystem quality and resource depletion but shows higher human health impacts. Scenario analysis reveals profitability at processing capacities over 100 000 kg·h–1 for PG with REE content above 0.5 wt %. However, more dilute sources (0.02–0.1 wt %) are not viable under current conditions due to acid and neutralization costs. This study offers the first in-depth sustainability assessment of REE recovery from PG waste and highlights key areas for future process development to improve access to low-grade sources and enhance environmental outcomes.

Keywords: rare earth elements (REE), techno-economic analysis (TEA), life cycle assessment (LCA), global uncertainty and sensitivity analysis, green chemical process design

graphic file with name es5c04952_0009.jpg

graphic file with name es5c04952_0008.jpg

1. Introduction

A major energy transition to clean energy is required to meet climate goals, which will require up to 700% more rare earth elements (REEs) to be produced. , Currently, approximately 90% of REE production occurs in China by mining REE containing ores (e.g., monazite, bastnaesite, and xenotime). This conventional REE production uses hydrometallurgical pathways to convert ores into saleable rare earth oxides (REO). , The process begins with mining the ores followed by beneficiation and leaching to extract the REEs from the other components. This concentrated REE stream is then separated by up to 300 stages of solvent extraction and refined into the final REO product which is sold to make other high value products (e.g., magnets). While this scheme has been profitable, it consumes large volumes of chemicals due to the intense acid/base leaching and the low separation factor of solvent extraction separations. This chemical consumption leads to negative environmental impacts from the production and transport of chemicals and the long-term storage of acidic tailings , (which consume lots of land and can fail catastrophically). The consumption of organic solvent and acid regenerant used in the solvent extraction has been shown to contribute up to 30% of the overall environmental impact of conventional REE production. This result motivates the development of new separation processes that use less energy and fewer toxic chemicals. One alternative technology that has been developed is solid-phase adsorption which uses ion-exchange or chelating resins to adsorb REEs from solution. Adsorption can be highly selective at low REE concentrations (such as those in dilute secondary sources) and has reduced chemical consumption compared to solvent extraction. However, resins are prohibitively expensive for their capacity and have low production rates compared to solvent extraction. , New biobased adsorbents that utilize proteins or peptides with high affinity and high separation factors between REEs could make solid-phase adsorption an attractive option for dilute REE recovery. , However, it is unclear whether this technology is economically feasible and more environmentally friendly than conventional techniques when implemented in systems for REE recovery.

Globally available secondary sources (e.g., product recycling, − coal fly ash, acid mine tailings) are potentially more sustainable alternatives to satisfy increasing REO demand. , Phosphogypsum (PG) is a potential secondary source that is a waste from fertilizer production (specifically phosphoric acid production). Currently, this PG waste is stored in large above ground stacks due to its classification as a toxic and radioactive waste. Within Florida alone, there are 200 million tons of PG stored in stacks which amounts to 1,000 tons of REEs. Considering the PG production rate in the U.S. (30 million tons PG·year–1), the annual consumption of REEs within the United States (9,000 tons REE·year–1) could be satisfied from PG alone assuming a minimum REE concentration of 0.02–0.03 wt % in PG. , Even higher PG REE concentrations (0.03–0.9 wt %) can be found around the world (e.g., Poland, Brazil, Russia), which may lead to more profitable operations. Alternative PG remediation systems (e.g., road construction) and other broader applications show some promise but are not profitable nor do they meet health and environmental standards. Therefore, the extraction of REEs from PG should be explored as a potentially profitable and environmentally friendly alternative to stack treatment. One study of a pilot-scale system for REE recovery from PG in Poland showed profitability at high risk and higher environmental impact than the conventional stack treatment. However, this study did not allocate any impacts to coproducts from the leaching operation. Further, this study had a limited scope considering only a mixed REE product (lower value compared to pure individual REOs after separation) and heavy reliance on fossil energy. Therefore, further work must rigorously explore process alternatives, especially in leaching, for sustainable REE recovery systems from PG.

Since the optimal process scheme for REE recovery from PG is still unknown, studies have primarily examined the removal of REEs from the PG crystal lattice. Many studies have accomplished this using recrystallization, carbonation, bioleaching, organic leaching, and most commonly, inorganic acid leaching. − These acid leaching experiments examined the effect of different parameters (e.g., leaching time, acid-to-solid ratio, temperature) and lixiviant (e.g., nitric acid, sulfuric acid, hydrochloric acid) on leaching efficiency. An advantage of using sulfuric acid is the coproduction of gypsum from the leaching process, but sulfuric acid has a lower REE leaching efficiency (approximately 40%) compared to hydrochloric acid and nitric acid (approximately 60%). ,, Presently, very little work has been done that applies this leaching data, in combination with a suitable selective separation, to evaluate the potential profitability and environmental impact of a REE recovery system. Despite the challenges, this work is especially rewarding due to the ease of making large process changes at this stage (low financial investment) and the ability to provide direction for future research for faster implementation. ,

2. Methods

2.1. System Overview

The system developed here, the recovery of rare earth elements from PG system (REEPS), achieves two objectives: 1) to remediate PG waste 2) to produce high purity (>99%) REOs. Figure a shows the system diagram used for the LCA broken down by process section including input and output flows. The complete process flow diagram (Figure b) shows each unit operation, input and output stream, and the phase of each stream. Each process section is organized by the same color shown in Figure a. The first section (leaching) takes the PG waste feedstock and extracts the REEs from the solid gypsum lattice to the liquid phase. Next, these REEs are separated from the leached solids and are concentrated using precipitation by oxalic acid. The mixed REE-oxalate precipitate is filtered and resuspended prior to the REE-specific bioadsorption (selective separation). The selective separation uses an REE selective biomolecule (e.g., Lanmodulin protein) attached to an agarose resin to create pure individual REE streams. In the refining section, each individual REE stream from the selective separation is then precipitated by oxalic acid, filtered, and calcined to the final REO product. The wastewater from the process is neutralized using sodium hydroxide and all heavy metals are precipitated after the addition of sodium phosphate to achieve high pH (around 8). The wastewater after this neutralization is sent to an onsite wastewater treatment facility (primary, secondary, and tertiary treatment) to remove organics, solids, and other ions to acceptable levels for release to the environment. Full details of the process model can be found in S1.

1.

1

Open in a new tab

(a) System diagram and (b) process flow diagram for the rare earth elements recovery from phosphogypsum system (REEPS).

2.2. Life Cycle Assessment (LCA)

To quantify the environmental impact of the REEPS system, a cradle-to-gate LCA was performed following ISO 14040/14044 and current industry standards. , Two functional units were used: 1 kg of PG remediated (to assess the sustainability footprint of waste valorization compared to PG stacking) and 1 kg of REO produced (to assess the sustainability footprint of REE production compared to conventional REO production).

The goal of the LCA study is to compare the new REEPS system to conventional approaches, identify the hotspots, and combine with global sensitivity analysis to determine the most influential process sections and parameters governing system-level sustainability. The system boundary (Figure a) includes impacts from raw material acquisition (cradle) through the production of the REO product (gate). A system expansion approach was used for coproduct handling. For the functional unit of PG remediated, the system was credited for the avoided production of gypsum and REOs. For the functional unit of REO produced, the system was credited for the avoided production of gypsum and the elimination of the PG waste. The system’s direct emissions to water and air are quantified in addition to flows from raw material and chemical production. To model background processes, activities from the ecoinvent v3.9.1 cutoff database were used. The US-SERC electricity grid was used to model the impact of electricity consumption due to the large volume of PG in this region. Impacts from construction and demolition were considered negligible and ignored. Transportation of PG to the system was not modeled in this study since there was no understanding of where the system should be built and what scale was optimal. In addition, the end of life for the concentrated radionuclide stream is unknown (e.g., storage, further refinement) and no impact was modeled for this flow.

To assess life cycle environmental impacts, the ReCiPe 2016 LCIA method (v1.03) was chosen to characterize the REEPS system impacts. ReCiPe 2016 was chosen because it is widely used and includes a variety of impact categories relevant to this system. Specifically, ReCiPe 2016 comprehensively accounts for impacts from radioactive substances, toxicity, and land use which are relevant for mined rock leaching wastes (e.g., PG and acid mining tailings). Since PG and mining tailings are wastes stored for long or indefinite periods, the “long-term” version of each LCIA method was used.

2.3. Techno-Economic Analysis (TEA)

A TEA was performed to assess the profitability of the system by three indicators (net present value at a 15% interest rate (NPV15), internal rate of return (IRR), and minimum selling price (MSP)). The TEA used a discounted cash flow rate of return analysis to assess profitability. To calculate the capital investment, the purchase costs of equipment were estimated using cost correlations. Changes in equipment cost between scenarios were adjusted using calculated size factors. , Equipment costs were converted from past dollar values to 2022 U.S. dollars using the chemical engineering plant cost index. From these equipment costs, the total capital investment was calculated using the Lang factor method (Lang factor of 4.28). Capital was depreciated using the MACRS 7-year depreciation schedule. A plant life of 30 years with an uptime of 90%, a tax rate of 28%, and an interest rate of 15% was used for the analysis. Plant construction was completed in a 3-year period with 8%, 60%, and 32% of the capital expended in each year, respectively. Working capital was estimated as 5% of the fixed capital investment.

Operating costs were calculated as the sum of variable and fixed operating costs. For variable operating costs, bulk chemical and utility prices were gathered from literature, government, and market sources (S3.1). The value of the REO product stream was estimated using a ‘basket price’ that considers the abundance of different REEs within PG (calculation in S3.2). The cost of the selective separation adsorbent was estimated as the sum of the price for specialized ion exchange resins and protein immobilized within this resin (calculation in S3.3). The cost of adsorbent replacements was annualized and included in the variable operating cost. The initial cost of the adsorbent was considered as the installed equipment cost for the selective separation (more discussion in S1.3.4). The producer price index was used to adjust prices to 2022 dollars. The fixed operating cost was calculated as the sum of labor, maintenance, and administrative costs.

2.4. Uncertainty Characterization and Global Sensitivity Analysis

The foreground inventory was compiled by modeling the entire process (Figure b) in Python v3.10.13 (available at 10.5281/zenodo.15126344). This code leverages several established packages (e.g., brightway2, bioSTEAM, , QSDsan) to combine the process modeling, LCA, TEA, and Monte Carlo global uncertainty and sensitivity analysis into one tool. The Monte Carlo analysis used Latin Hypercube sampling (3000 samples) to evaluate the uncertainty and sensitivity. Other numbers of samples (500 and 1000) were compared to ensure that results were reproduceable (SI). The sensitivity of parameters was calculated using Spearman’s rank correlation coefficients. The Spearman rank coefficients were calculated separately for technological and contextual parameters because technological parameters can be controlled by engineers, while contextual parameters are dictated by external forces. By separating these two types of parameters, it becomes clearer which controllable parameters are most influential and should be the focus of future research. Parameters, assigned parameter uncertainty distributions, and the corresponding references for assigning these distributions are provided in S2 and S3.1.

3. Results and Discussion

3.1. Baseline Results

3.1.1. Preoptimization of the System

Though there are several possible acids to use for leaching, we chose sulfuric acid for this analysis due to its prevalence in the literature, as well as its formation of a saleable coproduct, gypsum. Prior to completing the analysis, the global optimal values of leaching parameters for sulfuric acid were identified to ensure a more realistic estimate of profitability. The most profitable values of four key parameters were identified (bright yellow regions in Figure ) for sulfuric acid leaching of PG (leaching temperature of 47 °C, acid concentration 2.8 wt %, leaching time of 200 min, and liquid-to-solid ratio (L/S ratio) of 275 wt % liquid). The black dots on Figure represent the experimental local optimal conditions (experimental results provided S4), which are different than the global, system optimal conditions determined here. The experimental data was not extrapolated when finding the most profitable configuration. These optimal parameter values are only valid for this system since they represent the global optimal of the process. In addition, the limited data availability inhibited the consideration of interaction effects between parameters. Further details along with the complete set of the contour plots are available in S4.

2.

2

Open in a new tab

Contour plots showing how the values of key technological parameters were chosen to optimize net present value for the leaching unit. The black dot represents the experimentally determined local optimal conditions, which are different from the system’s global optimal conditions (the bright yellow regions).

3.1.2. Baseline TEA Results

In the baseline scenario, the REEPS system has a return on investment (ROI) of 23% and a payback period of 4 years (for a 30-year investment). Considering the time value of money using a discounted cash flow analysis, the system is profitable (NPV15 above zero and an IRR above 15%). The high NPV ($570 million) suggests that the system can be profitable. However, the IRR of 20% is a result of high upfront costs from capital expenditure and reduced returns over the system lifetime. Further, the REEPS system has a high level of uncertainty due to the low technological readiness of the system, indicating that this system could be a risky investment compared to other investments. As observed in the probability density plot (Figure ), the NPV15 ranges from around -$1 billion to $2 billion with the most probable value being around the baseline at $570 million. This uncertainty for the economic indicators is larger than what is observed for environmental indicators. This difference may be due to the uncertainty in environmental indicators being driven solely by technological parameters (e.g., leaching temperature), whereas economic indicators are subject to changes in both technological and contextual parameters (e.g., chemical prices, REO market prices, etc.). Increasing the technological readiness of technologies relevant to REE recovery from PG is key to reducing the large uncertainties in system sustainability.

5.

5

Open in a new tab

Baseline value (dots) and the probability density (shaded regions) of indicator values for the two functional units: (a) 1 kg of REO produced and (b) 1 kg of PG remediated. Peaks represent a higher probability of an indicator value, while broad distributions represent greater uncertainty. The indicator values of conventional REO production (dotted black line) are shown for comparison (with NPV15 > 0 indicating profitability). The indicator underlined with a solid black bar is an economic indicator (net present value at an interest rate of 15%) evaluated by TEA. All other indicators were evaluated by LCA using the ReCiPe 2016 LCIA method. The probability densities were calculated by collecting 3000 samples of the system using Latin Hypercube sampling from defined parameter distributions (S2 and S3.1). The full figure with all environmental and economic indicators can be found in S5.

Though the REEPS system can be profitable, it is important to further understand what costs are driving the profitability to guide further research and process improvements. Figure b shows the breakdown of the costs contributing to the MSP ($44·kg–1·REO) by process section. Approximately half of the MSP is from operating costs ($23·kg–1·REO, composed of chemicals and utilities) with the other half being related to capital recovery cost ($20·kg–1·REO composed of capital depreciation, average income tax, and average return on investment). In addition, there is a smaller contribution due to fixed operating costs ($5.2·kg–1·REO) and a credit from the sale of the byproduct gypsum (−$4.3·kg–1·REO). The majority of the capital cost is from the bio adsorption resin and the biomolecule ligand ($730 million and $240 million, respectively) in the selective separation ($15·kg–1·REO). Replacements for the adsorbent comprise the majority of the chemical and material costs for the selective separation as well. Comparing to the conventional solvent extraction approach ($18–97·kg–1·REO), , the biobased selective separation and refining has similar cost ($24·kg–1·REO). The concentration and refining section costs ($2.9·kg–1·REO and $1.2·kg–1·REO, respectively) are primarily due to precipitant consumption, specifically oxalic acid. The leaching section costs ($4.1·kg–1·REO) are split between capital, utilities, and chemicals ($8.3·kg–1·REO) and the gypsum byproduct (-$4.3·kg–1·REO). The chemical cost of leaching is primarily from sulfuric acid while the primary utility cost is for natural gas heating. The wastewater treatment section expenses ($12·kg–1·REO) are split between operating costs ($9.6·kg–1·REO) and capital costs ($2.0·kg–1·REO). The primary operating cost is the sodium hydroxide for neutralizing the acid waste from the leaching section. Therefore, reducing acid use and selective separation costs should be a priority for improving the profitability of the system.

3.

3

Open in a new tab

(a) Environmental impact of the REEPS system compared to conventional REO production and PG stack treatment using the ReCiPe 2016 LCIA method. The relative impact of the REEPS system compared to conventional techniques shows the extent of REEPS advantages (blue text) and disadvantages (red text). Midpoint category impacts are shown above the dotted line with end point impacts below the dotted line. (b) The contribution of each process section toward the minimum selling price of the REO product.

3.1.3. Baseline LCA Results

When considering the functional unit of one kg of REO produced, the REEPS system has a mixed performance compared to conventional REO production (Figure a). We used both end point and midpoint impact assessment to understand system sustainability compared to conventional methods. End point impact methods use value weighting to condense multiple midpoint impact categories (e.g., acidification and ecotoxicity) into one broad end point impact category (e.g., ecosystem quality). Therefore, end point analysis is helpful to reduce the complexity of analyzing trade-offs but at the cost of detail and accuracy. The end point analysis shows that the REEPS system outperforms conventional REO production in ecosystem quality (93.0% of the impact) and resource depletion (96.2% of the impact) but underperforms in human health (213% of the impact). These advantages of the REEPS system are led by reductions in land use, eutrophication, ecotoxicity, human toxicity, and ionizing radiation. The advantages in ionizing radiation, human toxicity, and land use are largely due to the avoided impact from the stacking of PG waste (shown in Figure ). Advantages in the other categories result from several high impact processes in conventional REO production that are not present in the REEPS system: storage of large volumes of acid mining tailings which can leach radioisotopes and toxic substances into the environment and toxic solvent use in the selective separation which contributes up to 30% of the impact of conventional REO processing.

4.

4

Open in a new tab

Contribution of each process section toward the environmental and economic impacts of the system using a functional unit of (a) 1 kg of REO produced and (b) 1 kg of PG remediated.

Conversely, the REEPS system has higher impact (150–400%) on climate change, particulate matter formation, acidification, water use, and fossil depletion. Utilities account for 0.21% (ozone depletion) to 18% (energy resources) of the impact for the impact across selected categories. Overall, chemical consumption is the primary contributor to the environmental impacts of the system. For acidification, approximately 70% of the impact results from the consumption of sulfuric acid in leaching. Fossil depletion impact is largely due to the consumption of sodium hydroxide (73%) used for neutralization of acidic wastewater. Climate change, particulate matter formation, and acidification impacts are more evenly attributed to various chemicals used in the process (Figure ). The REEPS photochemical oxidant formation and ozone depletion impacts are much higher than conventional REO production (approximately 940% and 1450% of the impact, respectively). These high impacts are almost exclusively due to oxalic acid consumption (approximately 92% and 83% of the impact, respectively). Therefore, future process modifications that reduce (or eliminate) the consumption of these chemicals, like oxalic acid, can greatly reduce the environmental impact of the REEPS system. Additionally, we compared impacts from REEPS without wastewater treatment (REEPS w/o WWT) to conventional REO production (cREO), since cREO systems mostly exclude wastewater treatment from the system scope. Notably, the climate change impact for REEPS w/o WWT and cREO were very similar (80 and 77 kg CO2 eq·kg–1·REO, respectively). The full results are shown in Figure a.

When comparing to conventional PG stack treatment (functional unit of 1 kg of PG remediated), the REEPS system again has mixed performance. End point analysis shows that the REEPS system has lower impacts than the PG stack treatment in ecosystem quality (97.5% of the impact) and resource depletion (−2.46 and 0 USD 2013·kg–1·PG remediated, respectively), but higher impact on human health (1380% of impact). The main contributors to each impact category are the same as described above (the system operates identically regardless of which functional unit is considered). However, when comparing to the PG stack system, the avoided production of REOs from the conventional route is credited to the system. This credit is about the same magnitude as the impacts of the REEPS system for half the impact categories (Figure ). For marine eutrophication, land use, and material resources, the credit is over 90% of the impact leading to large net negative impact for the REEPS system. However, for other categories (e.g., ozone depletion and photochemical oxidant formation), the credit is minimal (<10%) further emphasizing the need for process alternatives for the concentration section.

3.1.4. Hotspot Analysis

Three process sections (leaching, concentration, and wastewater treatment) are responsible for over 75% of the environmental impacts (Figure ). The primary impact from those process sections is from chemical consumption. The leaching section uses sulfuric acid in large volumes to extract the REEs from the solid gypsum matrix at elevated temperatures generated by burning natural gas. Though, a modest benefit is observed for the generation of a saleable gypsum coproduct (especially for particulate formation and land transformation). The concentration section requires a large excess of oxalic acid to bind metals in the leachate and to precipitate the REEs from the solution in preparation for the selective separation. The wastewater treatment section neutralizes the acidic wastewater from leaching and precipitates the rest of the toxic heavy metals with sodium hydroxide and sodium phosphate. Conventional REO production has a higher proportion of environmental impacts from the selective separation since solvent extraction has low selectivity between REEs requiring many stages and complex operations. , However, the selective separation in the REEPS system cannot be fairly compared since the model does not include the chemicals for regeneration or pumping costs (insufficient data). Though these impacts are expected to be minimal compared to solvent extraction. , The REEPS system also does not include further refining processes like electrolysis which contribute roughly 10–30% of the total impact of conventional REO production. In this study it was assumed that the separation produced high enough purity REOs that further refining would not be necessary, but this will be important to revisit in future work using a more detailed separation model.

To avoid the impact of the most influential chemicals, changes to the REEPS process scheme should be explored. Specifically, improvements (or alternatives) to acid leaching for moving REEs from the solid phase to the liquid phase could reduce system impacts by up to 60%. Improvements in leaching could come from using acids with higher REE leaching efficiency and understanding, fundamentally, how acids and technical parameters influence this leaching efficiency. However, acid leaching not only reduces sustainability through chemical consumption and inefficient extraction, but also by requiring neutralization of the acidic wastewater now contaminated with toxic heavy metals. To mitigate this downstream problem, other neutralizing agents (e.g., magnesium carbonate, ammonium hydroxide) could be examined for potential trade-offs between profitability and environmental impact. Additionally, alternative ‘leaching’ techniques that do not require large volumes of acid and have high efficiency (e.g., ammono-carbonation) ,, are extremely promising. In addition, improved technologies for the concentration section could lead to an additional 20% reduction in impact. Due to the dilute nature of the feedstock, we included the concentration section to isolate and concentrate the REEs from the leachate to make the downstream separation more effective. However, it is currently unclear what specifications the concentration section must meet to enable an efficient selective separation. By developing detailed models for the selective separation, we can clarify the requirements of the concentration section and further optimize the system. Depending on the required specifications, we can also consider alternative concentration technologies that reduce chemical consumption (e.g., electrodialysis, filtration) to identify more sustainable process schemes. Similarly, we can examine other selective separations (e.g., membrane adsorption, solvent extraction, nanofiltration). By expanding the established Python framework (10.5281/zenodo.15126344), we can ‘plug and play’ these different technologies to find more optimal process schemes and to examine trade-offs between technologies.

3.2. Probabilistic Sustainability

Sustainability assessments inherently involve uncertainties, stemming from variability in raw material inputs, process efficiencies, market fluctuations, and evolving regulatory frameworks. Traditional life cycle assessment (LCA) and techno-economic analysis (TEA) often rely on deterministic modeling, where fixed parameter values lead to point estimates that may not fully capture variability, particularly for early stage technologies. Since the aim of this work is to guide decision-making across different REE recovery technologies, we focus on the foreground uncertainties across the technological space. This will help stakeholders compare technologies and set appropriate R&D targets.

Here, we introduce the concept of “Probabilistic Sustainability” for assessing and evaluating the potential of early stage/emerging technologiesespecially for low technological readiness levels (TRL). , A Probabilistic Sustainability Assessment framework is proposed, we incorporate system preoptimization, uncertainty quantification and stochastic modeling to enhance decision-making robustness, making sustainability assessments more reflective of uncertainties in future real-world implementation. More importantly, this approach avoids false precision in decision-making and provides critical insights for policymakers to support more informed and resilient sustainability strategies.

3.2.1. Uncertainty Analysis

Since no large-scale system exists for REE recovery from PG, we first needed to identify the possible design space of the system. We defined the feedstock PG REE content, 0.5 wt %, by taking the average REE content of PG stacks globally (ranging from 0.02 to 0.9 wt % REE). Next, we used this REE content to calculate a reasonable plant capacity (approximately 1 M kg PG processed·hr–1). We ensured that the annual REO production rate was less than the global demand for REOs. Further, we confirmed that a PG containing region has a PG supply that could meet this REO production rate (S1.2). Additionally, we considered that some fraction of the REEs within PG would be unrecoverable using the current process scheme (due to the highly dilute nature of the feed source). Therefore, REEs with relative abundance below 1 wt % (Sm, Tb, Eu, Ho, Yb, Lu, Y, and Sc) were not modeled as part of the REE stream for the analysis. This consideration led to a loss in REE product of 2.2 wt % and reduced the value of the combined REO product from $55.0·kg–1·REO to $51.5·kg–1·REO. Using the above REE content, plant capacity, and recoverability, we defined a baseline scenario for REE recovery from PG. The baseline scenario results for 10 key sustainability indicators are shown in Figure . A figure including all indicators for each functional unit is provided in S5.

3.2.2. Global Sensitivity Analysis

We used a global sensitivity analysis to identify key parameters for process improvement (Figure ). The list of parameters along with their uncertainty distributions are provided in S2 and S3.1. A figure showing the sensitivity for the other functional unit is in S4. Of the technological parameters (Figure a), seven parameters had coefficients larger than 0.15: REE recovery of the selective separation (S1), sodium hydroxide required for neutralizing the wastewater (P3), oxalic acid for precipitation of REEs (P1 and P2), the concentration of acid for leaching (U1), the leaching overflow to underflow ratio (U1), and the leaching lixiviant to solid ratio (U1). The process unit operation codes are given in Figure b. One of the main drivers for the sensitivity of these parameters is their influence on the production of REOs. The REO product is very valuable and has a large environmental impact (avoided production credit) leading to increased sustainability as more REOs are produced. Therefore, minimizing waste from the selective separation (i.e., REE recovery) should be prioritized when designing the unit. Similarly, leaching efficiency should be prioritized in the leaching unit. Developing a greater understanding of what acids and parameters (e.g., acid concentration and acid-to-solid ratio) are most effective should be prioritized. Further, we need a better understanding of how scaling up bench-scale leaching experiments influences leaching efficiency (batch leaching at lab scale is unlikely to perform as well as large countercurrent flow systems that optimize thermodynamic driving forces). Understanding how operational parameters in these larger flow systems (e.g., overflow to underflow ratio) will also be important in reducing uncertainty in the process design. Another main driver for the sensitivity of the seven parameters is raw material consumption. Four of the parameters directly relate to material consumption: oxalic acid consumption in P1 and P2, sodium hydroxide consumption in P3, and acid concentration in U1. Therefore, using less of these chemicals, different chemicals, or different technologies altogether would lead to significant decreases in environmental impact and cost.

6.

6

Open in a new tab

The sensitivity of environmental and economic indicators to (a) technological parameters and (b) contextual parameters. Blue indicates better performance (reduced environmental impact and higher profitability) with an increase in parameter value, while red indicates inferior performance as the parameter increases. The more vibrant the color, the greater the sensitivity of an indicator (y-axis) to a change in a parameter (x-axis). The first indicator is an economic indicator (evaluated by TEA), and indicators below are environmental indicators (evaluated by LCA for the functional unit of 1 kg of PG remediated). Sensitivity was calculated using Spearman rank correlations with Latin Hypercube sampling (3000 samples) of parameter distributions (given in S2 and S3.1).

Regarding contextual parameters (Figure b), only economic indicators are relevant since changes in prices and TEA parameters do not affect environmental indicators. Interest rate and REO price are used to calculate IRR and MSP, respectively. Therefore, there is no sensitivity of these indicators to their respective parameters. Five contextual parameters have coefficients above 0.15: income tax rate, interest rate, sodium hydroxide price, REO price, biomolecule price, and number of operating days per year. Of these parameters, the interest rate and REO price are most influential to profitability. The interest rate is influential due to the high capital expenditure in early years making this investment riskier. As observed in Figure , reduction in the cost of the adsorbent should be prioritized to reduce sensitivity to this parameter by reducing capital cost. The REO price is also influential due to the wide range given to the distribution (36.1–67.0 $·kg–1·REO). Within the past decade REO prices have been highly volatile. Since the current REO supply chain still exists in a similar form, we made the distribution reflect this volatility. In addition to this volatility in REO prices, we also considered that increased demand for the green energy transition may lead to higher REO prices in the future (reflected by the positive correlation to profitability in Figure b). However, we acknowledge the possibility of REO prices decreasing. Since REOs are produced together but have different market demand, it means that REOs with the least demand will be produced in excess (decreasing the price). Similar to the large uncertainty in REO prices, the price of producing bulk biomolecules is highly uncertain. Some studies report theoretical bulk prices of proteins and peptides. However, more work is required to understand how much prices could vary for different molecules. The uncertainty range in this study ranges from the theoretical cost of bulk protein ($0.004·g–1) to bulk peptide ($10·g–1) with a more average number ($0.5·g–1) chosen for the base case value.

3.2.3. Scenario Analysis

We used a scenario analysis to explore which combination of plant capacity and PG REE content enable a profitable operation (Figure ). In general, profitability increases (shown by decreasing MSP) as capacity increases. This increase in profitability with capacity indicates that larger centralized facilities should be prioritized over decentralized modular systems. However, a centralized facility would require transportation of PG waste to the processing plant. The implications of transportation on the sustainability of the system are not considered in this work. Future work will need to consider the logistics, cost, and environmental impact of transporting PG and identify regions that are ideal for a REE recovery facility. Regions that have large volumes of PG in close proximity (e.g., Florida with >1 billion tons in 24 stacks) are promising for investigation.

7.

7

Open in a new tab

Effect of capacity and REE content in the PG on process profitability (measured by minimum selling price (MSP)) (a) for the baseline system and (b) considering improvements in the separation technology through increased capacity (from 0.0058 to 0.025 mol REE·L-1 resin) and REE recovery (from 99 to 100%). Brighter regions indicate more profitable scenarios (lower MSP) where all contours to the right of the dotted lines are profitable. The dotted lines are the current basket REO selling price ($51.5·kg–1·REO) and a future scenario with two times the current selling price ($103·kg–1·REO). The REE content of PG stacks around the globe varies between 0.02 and 0.9 wt % REE.22.

However, the current REEPS system is only profitable for PG stacks with an REE content above roughly 0.5 wt %. To determine conditions that make operations with more dilute feedstocks profitable, we considered a few scenarios. First, we examined the effect of increased sales. If the value of REOs increases in the future (e.g., REO prices double), PG stacks with as little as 0.2 wt % REE content become profitable (Figure a). Another form of additional income could come from the producers or managers of PG waste. They may prefer to pay for remediation as opposed to having to manage a PG stack indefinitely with its risk of release. Second, we considered a reduction in capital cost. Even with improvements in the capacity of the separation unit (0.00577 to 0.025 mol·L–1 adsorbent), the most dilute PG stacks remain inaccessible (Figure b). These results indicate that a new process scheme must be developed to be able to access the most dilute PG stacks (0.02–0.1 wt %). New process schemes should reduce system costs by utilizing new technologies and address the most sensitive parameters. Specifically, the cost of acid for REE leaching and subsequent neutralization of the wastewater is the primary barrier to recovery from dilute sources. A technology that disrupts this paradigm would greatly increase the viability of REE recovery from dilute minerals.

3.3. Limitations and Path Forward

There are several limitations to the results of this study, which are consistent with those typically observed in critical mineral recovery sustainability analysis work and are inherent to the current methodologies commonly used in this field.

First, more reliable, relevant, and complete experimental data would increase certainty in the REEPS system results. For example, this work does not model the selective separation using thermodynamic binding constants or kinetic rate constants. We calculated profitability under uncertainty based on literature adsorbent capacity and assumptions for adsorbent lifetime and REE recovery (S1.3.4 and Table S8). As data becomes available, future work should develop a detailed model to reduce uncertainty and more accurately assess how process performance depends on thermodynamics and kinetics, as well as adsorbent lifetime. There are similar data limitations in the other process sections as well. Considering process alternatives for these technologies must be explored. Further, a better understanding of the life cycle impacts of conventional PG stack treatment and REO production is required. Many of the technical details of conventional REE production are not public information. , The illegal (and therefore unregulated) production of REEs, often occurring across various countries outside of US, makes it difficult to assess the full environmental impact of the current REO supply chain. A qualitative discussion of the limitations and uncertainty of LCAs for REO production is available in the literature, but uncertainty has not been rigorously quantified to date. ,

Further, studies of REE systems report results in terms of either a mixed REE product, a purified individual REO product (either as an individual REO like Nd or in total), or a refined rare earth metal product. These three different purity REE products, with greatly different market values, are a result of different system scopes making comparison between studies challenging. Due to uncertainty in results from other conventional REO studies (e.g., functional unit, allocation, scope), we compared environmental impacts to ecoinvent. However, even the ecoinvent results are not entirely transparent with their allocation. Ecoinvent performs mass-based subdivision of the system prior to economic allocation, but they do not describe how they choose which REOs are reference products (for subdivision) and which are allocateable byproducts (for economic allocation). Therefore, we converted the ecoinvent results from a functional unit of individual separated REOs (e.g., Nd2O3) to total separated REOs for comparison to the total amount of separated REOs considered here. We do acknowledge that a mixture of separated REOs has different value due to its composition. However, we recommend using this total separated REO functional unit for comparison to avoid uncertainties due to allocation, especially for feedstocks of similar composition like we have here.

For PG stack treatment, only two studies have quantified the environmental impact. , Of these, only one provided a detailed life cycle inventory. These two studies disagreed on the primary impact of PG (e.g., respiratory inorganics is the highest impact in one study and has no impact in the other) suggesting we do not understand what flows to consider and how different geographies affect LCA results. The LCA methodology is not effective at quantifying the long-term impacts of stored wastes on the environment. Decisions on how to treat long-term impacts from mining have been shown to change impact by up to 8 orders of magnitude. This limitation must be resolved moving forward as it restricts our ability to consider temporally relevant impacts to accurately assess the sustainability of technologies. ,

Future efforts in LCIA methodologies in this field should also improve modeling methods that quantify the impacts of radioactivity and water depletion. For radioactivity, current LCIA methods calculate the direct impact of radioactive substances on human health but do not consider the indirect impact of these radionuclides on ecosystems. In addition, current LCIA methods do not have a way to reliably link impacts to how radionuclide concentrations change as they move through the environment. Some models outside of the LCA field exist that could be integrated into LCIA methodologies to more accurately quantify the impact of radionuclide releases on organism and ecosystem health. For impacts of water depletion, most LCIA methods consider only water flows that do not return to the native aquifer. However, water reclamation systems, like REEPS, return water back to the native aquifer as part of their function. Therefore, it is important to note that the water use impact used here, along with those reported in other current REE recovery studies, is not comprehensive. The REEPS system could theoretically have a net negative ‘water depletion’ impact since water is being returned to the environment.

In summary, this work will enable the community to standardize sustainability assessments and purification concepts of REE recovery and prioritize R&D pathways. We established open-access modeling tools to translate REE separation concepts to field-scale systems and evaluated the potential economic and environmental feasibility under uncertainty.

Supplementary Material

es5c04952_si_001.pdf (928.8KB, pdf)

Acknowledgments

We thank the National Science Foundation for their support under Award Number ECO-CBET-2133530.

The code used in the analysis is available at 10.5281/zenodo.15126344.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.est.5c04952.

  • Additional text, figures, and tables that contextualize and extend the analysis described in the main text (PDF)

The authors declare no competing financial interest.

References

  1. IPCC . Global Warming of 1.5°C. An IPCC Special Report on the Impacts of Global Warming of 1.5°C above Pre-Industrial Levels and Related Global Greenhouse Gas Emission Pathways, in the Context of Strengthening the Global Response to the Threat of Climate Change, Sustainable Development, and Efforts to Eradicate Poverty; Cambridge University Press, 2022; 10.1017/9781009157940. [DOI] [Google Scholar]
  2. Role of Critical Minerals in Clean Energy Transitions. IEA: Paris, 2021; https://www.iea.org/reports/the-role-of-critical-minerals-in-clean-energy-transitions. [Google Scholar]
  3. Gupta C. K., Krishnamurthy N.. Extractive Metallurgy of Rare Earths. International Materials Reviews. 1992;37(1):197–248. doi: 10.1179/imr.1992.37.1.197. [DOI] [Google Scholar]
  4. Shahbaz A.. A Systematic Review on Leaching of Rare Earth Metals from Primary and Secondary Sources. Minerals Engineering. 2022;184:107632. doi: 10.1016/j.mineng.2022.107632. [DOI] [Google Scholar]
  5. Vahidi E., Zhao F.. Environmental Life Cycle Assessment on the Separation of Rare Earth Oxides through Solvent Extraction. Journal of Environmental Management. 2017;203:255–263. doi: 10.1016/j.jenvman.2017.07.076. [DOI] [PubMed] [Google Scholar]
  6. Zapp P., Schreiber A., Marx J., Kuckshinrichs W.. Environmental Impacts of Rare Earth Production. MRS Bull. 2022;47(3):267–275. doi: 10.1557/s43577-022-00286-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Schreiber A., Marx J., Zapp P.. Life Cycle Assessment Studies of Rare Earths Production - Findings from a Systematic Review. Science of The Total Environment. 2021;791:148257. doi: 10.1016/j.scitotenv.2021.148257. [DOI] [PubMed] [Google Scholar]
  8. Opare E. O., Struhs E., Mirkouei A.. A Comparative State-of-Technology Review and Future Directions for Rare Earth Element Separation. Renewable and Sustainable Energy Reviews. 2021;143:110917. doi: 10.1016/j.rser.2021.110917. [DOI] [Google Scholar]
  9. El Ouardi Y., Virolainen S., Massima Mouele E. S., Laatikainen M., Repo E., Laatikainen K.. The Recent Progress of Ion Exchange for the Separation of Rare Earths from Secondary Resources - A Review. Hydrometallurgy. 2023;218:106047. doi: 10.1016/j.hydromet.2023.106047. [DOI] [Google Scholar]
  10. Dong Z., Mattocks J. A., Deblonde G. J.-P., Hu D., Jiao Y., Cotruvo J. A. Jr., Park D. M.. Bridging Hydrometallurgy and Biochemistry: A Protein-Based Process for Recovery and Separation of Rare Earth Elements. ACS Cent. Sci. 2021;7(11):1798–1808. doi: 10.1021/acscentsci.1c00724. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Dong Z., Mattocks J. A., Seidel J. A., Cotruvo J. A., Park D. M.. Protein-Based Approach for High-Purity Sc, Y, and Grouped Lanthanide Separation. Sep. Purif. Technol. 2024;333:125919. doi: 10.1016/j.seppur.2023.125919. [DOI] [Google Scholar]
  12. Li Z., Diaz L. A., Yang Z., Jin H., Lister T. E., Vahidi E., Zhao F.. Comparative Life Cycle Analysis for Value Recovery of Precious Metals and Rare Earth Elements from Electronic Waste. Resources, Conservation and Recycling. 2019;149:20–30. doi: 10.1016/j.resconrec.2019.05.025. [DOI] [Google Scholar]
  13. Azimi G., Sauber M. E., Zhang J.. Technoeconomic Analysis of Supercritical Fluid Extraction Process for Recycling Rare Earth Elements from Neodymium Iron Boron Magnets and Fluorescent Lamp Phosphors. Journal of Cleaner Production. 2023;422:138526. doi: 10.1016/j.jclepro.2023.138526. [DOI] [Google Scholar]
  14. Chowdhury N. A., Deng S., Jin H., Prodius D., Sutherland J. W., Nlebedim I. C.. Sustainable Recycling of Rare-Earth Elements from NdFeB Magnet Swarf: Techno-Economic and Environmental Perspectives. ACS Sustainable Chem. Eng. 2021;9(47):15915–15924. doi: 10.1021/acssuschemeng.1c05965. [DOI] [Google Scholar]
  15. Alipanah M., Park D. M., Middleton A., Dong Z., Hsu-Kim H., Jiao Y., Jin H.. Techno-Economic and Life Cycle Assessments for Sustainable Rare Earth Recovery from Coal Byproducts Using Biosorption. ACS Sustainable Chem. Eng. 2020;8(49):17914–17922. doi: 10.1021/acssuschemeng.0c04415. [DOI] [Google Scholar]
  16. Fritz A. G., Tarka T. J., Mauter M. S.. Technoeconomic Assessment of a Sequential Step-Leaching Process for Rare Earth Element Extraction from Acid Mine Drainage Precipitates. ACS Sustainable Chem. Eng. 2021;9(28):9308–9316. doi: 10.1021/acssuschemeng.1c02069. [DOI] [Google Scholar]
  17. Fritz A. G., Tarka T. J., Mauter M. S.. Assessing the Economic Viability of Unconventional Rare Earth Element Feedstocks. Nat. Sustain. 2023;6(9):1103–1112. doi: 10.1038/s41893-023-01145-1. [DOI] [Google Scholar]
  18. Jyothi R. K., Thenepalli T., Ahn J. W., Parhi P. K., Chung K. W., Lee J.-Y.. Review of Rare Earth Elements Recovery from Secondary Resources for Clean Energy Technologies: Grand Opportunities to Create Wealth from Waste. Journal of Cleaner Production. 2020;267:122048. doi: 10.1016/j.jclepro.2020.122048. [DOI] [Google Scholar]
  19. US EPA, O. TENORM: Fertilizer and Fertilizer Production Wastes. https://www.epa.gov/radiation/tenorm-fertilizer-and-fertilizer-production-wastes (accessed Oct 26, 2022).
  20. Phosphogypsum Stacks. https://fipr.floridapoly.edu/about-us/phosphate-primer/phosphogypsum-stacks.php (accessed Dec 21, 2023).
  21. U.S.: rare earths consumption 2022. Statista. https://www.statista.com/statistics/616726/consumption-of-rare-earths-in-the-united-states/ (accessed Jan 11, 2024).
  22. Mukaba J.-L., Eze C. P., Pereao O., Petrik L. F.. Rare Earths’ Recovery from Phosphogypsum: An Overview on Direct and Indirect Leaching Techniques. Minerals. 2021;11(10):1051. doi: 10.3390/min11101051. [DOI] [Google Scholar]
  23. Liang H., Zhang P., Jin Z., DePaoli D.. Rare Earths Recovery and Gypsum Upgrade from Florida Phosphogypsum. Mining, Metallurgy & Exploration. 2017;34(4):201–206. doi: 10.19150/mmp.7860. [DOI] [Google Scholar]
  24. Tsioka M., Voudrias E. A.. Comparison of Alternative Management Methods for Phosphogypsum Waste Using Life Cycle Analysis. Journal of Cleaner Production. 2020;266:121386. doi: 10.1016/j.jclepro.2020.121386. [DOI] [Google Scholar]
  25. Kulczycka J., Kowalski Z., Smol M., Wirth H.. Evaluation of the Recovery of Rare Earth Elements (REE) from Phosphogypsum Waste - Case Study of the WIZÓW Chemical Plant (Poland) Journal of Cleaner Production. 2016;113:345–354. doi: 10.1016/j.jclepro.2015.11.039. [DOI] [Google Scholar]
  26. Rychkov V., Kirillov E., Kirillov S., Bunkov G., Botalov M., Semenishchev V., Smyshlyaev D., Malyshev A., Taukin A., Akcil A.. Rare Earth Element Preconcentration from Various Primary and Secondary Sources by Polymeric Ion Exchange Resins. Separation & Purification Reviews. 2022;51:468. doi: 10.1080/15422119.2021.1993255. [DOI] [Google Scholar]
  27. Salo M., Knauf O., Mäkinen J., Yang X., Koukkari P.. Integrated Acid Leaching and Biological Sulfate Reduction of Phosphogypsum for REE Recovery. Minerals Engineering. 2020;155:106408. doi: 10.1016/j.mineng.2020.106408. [DOI] [Google Scholar]
  28. Thompson V. S., Gupta M., Jin H., Vahidi E., Yim M., Jindra M. A., Nguyen V., Fujita Y., Sutherland J. W., Jiao Y., Reed D. W.. Techno-Economic and Life Cycle Analysis for Bioleaching Rare-Earth Elements from Waste Materials. ACS Sustainable Chem. Eng. 2018;6(2):1602–1609. doi: 10.1021/acssuschemeng.7b02771. [DOI] [Google Scholar]
  29. Habashi F.. The Recovery of the Lanthanides from Phosphate Rock. Journal of Chemical Technology and Biotechnology. Chemical Technology. 1985;35(1):5–14. doi: 10.1002/jctb.5040350103. [DOI] [Google Scholar]
  30. Jarosiński A., Kowalczyk J., Mazanek Cz.. Development of the Polish Wasteless Technology of Apatite Phosphogypsum Utilization with Recovery of Rare Earths. J. Alloys Compd. 1993;200(1):147–150. doi: 10.1016/0925-8388(93)90485-6. [DOI] [Google Scholar]
  31. Walawalkar M., Nichol C. K., Azimi G.. Process Investigation of the Acid Leaching of Rare Earth Elements from Phosphogypsum Using HCl, HNO3, and H2SO4. Hydrometallurgy. 2016;166:195–204. doi: 10.1016/j.hydromet.2016.06.008. [DOI] [Google Scholar]
  32. Cánovas C. R., Chapron S., Arrachart G., Pellet-Rostaing S.. Leaching of Rare Earth Elements (REEs) and Impurities from Phosphogypsum: A Preliminary Insight for Further Recovery of Critical Raw Materials. Journal of Cleaner Production. 2019;219:225–235. doi: 10.1016/j.jclepro.2019.02.104. [DOI] [Google Scholar]
  33. Moni S. M., Mahmud R., High K., Carbajales-Dale M.. Life Cycle Assessment of Emerging Technologies: A Review. Journal of Industrial Ecology. 2020;24(1):52–63. doi: 10.1111/jiec.12965. [DOI] [Google Scholar]
  34. Li Y., Trimmer J. T., Hand S., Zhang X., Chambers K. G., Lohman H. A. C., Shi R., Byrne D. M., Cook S. M., Guest J. S.. Quantitative Sustainable Design (QSD) for the Prioritization of Research, Development, and Deployment of Technologies: A Tutorial and Review. Environmental Science: Water Research & Technology. 2022;8:2439. doi: 10.1039/D2EW00431C. [DOI] [Google Scholar]
  35. ISO 14040:2006; Environmental management: Life cycle assessment; Principles and framework; International Organization for Standardization, 2006. https://www.iso.org/standard/37456.html (accessed May 27, 2022).
  36. ISO 14044:2006; Environmental management: Life cycle assessment; Requirements and guidelines; International Organization for Standardization, 2006. https://www.iso.org/standard/38498.html (accessed May 27, 2022).
  37. Wernet G., Bauer C., Steubing B., Reinhard J., Moreno-Ruiz E., Weidema B.. The Ecoinvent Database Version 3 (Part I): Overview and Methodology. Int. J. Life Cycle Assess. 2016;21(9):1218–1230. doi: 10.1007/s11367-016-1087-8. [DOI] [Google Scholar]
  38. Huijbregts M. A. J., Steinmann Z. J. N., Elshout P. M. F., Stam G., Verones F., Vieira M., Zijp M., Hollander A., van Zelm R.. ReCiPe2016: A Harmonised Life Cycle Impact Assessment Method at Midpoint and Endpoint Level. Int. J. Life Cycle Assess. 2017;22(2):138–147. doi: 10.1007/s11367-016-1246-y. [DOI] [Google Scholar]
  39. Seider, W. D. ; Seader, J. D. ; Lewin, D. R. . Product and Process Design Principles: Synthesis, Analysis and Design, 3rd ed.; John Wiley & Sons, Ltd., 2009. [Google Scholar]
  40. Perry, R. H. ; Green, D. W. . Perry’s Chemical Engineers’ Platinum Edition, 7th ed.; McGraw-Hill, 1999. [Google Scholar]
  41. Tan, E. C. D. ; Talmadge, M. ; Dutta, A. ; Hensley, J. ; Schaidle, J. ; Biddy, M. ; Humbird, D. ; Snowden-Swan, L. J. ; Ross, J. ; Sexton, D. ; Yap, R. ; Lukas, J. . Process Design and Economics for the Conversion of Lignocellulosic Biomass to Hydrocarbons via Indirect Liquefaction. Thermochemical Research Pathway to High-Octane Gasoline Blendstock Through Methanol/Dimethyl Ether Intermediates. NREL/TP-5100-62402; National Renewable Energy Laboratory (NREL): Golden, CO, United States, 2015; 10.2172/1215006. [DOI] [Google Scholar]
  42. Mutel C.. Brightway: An Open Source Framework for Life Cycle Assessment. Journal of Open Source Software. 2017;2(12):236. doi: 10.21105/joss.00236. [DOI] [Google Scholar]
  43. Cortes-Peña Y., Kumar D., Singh V., Guest J. S.. BioSTEAM: A Fast and Flexible Platform for the Design, Simulation, and Techno-Economic Analysis of Biorefineries under Uncertainty. ACS Sustainable Chem. Eng. 2020;8(8):3302–3310. doi: 10.1021/acssuschemeng.9b07040. [DOI] [Google Scholar]
  44. Shi R., Guest J. S.. BioSTEAM-LCA: An Integrated Modeling Framework for Agile Life Cycle Assessment of Biorefineries under Uncertainty. ACS Sustainable Chem. Eng. 2020;8(51):18903–18914. doi: 10.1021/acssuschemeng.0c05998. [DOI] [Google Scholar]
  45. Li Y., Zhang X., Morgan V. L., Lohman H. A. C., Rowles L. S., Mittal S., Kogler A., Cusick R. D., Tarpeh W. A., Guest J. S.. QSDsan: An Integrated Platform for Quantitative Sustainable Design of Sanitation and Resource Recovery Systems. Environ. Sci.: Water Res. Technol. 2022;8:2289. doi: 10.1039/D2EW00455K. [DOI] [Google Scholar]
  46. Groen E. A., Bokkers E. A. M., Heijungs R., de Boer I. J. M.. Methods for Global Sensitivity Analysis in Life Cycle Assessment. Int. J. Life Cycle Assess. 2017;22(7):1125–1137. doi: 10.1007/s11367-016-1217-3. [DOI] [Google Scholar]
  47. Das S., Gaustad G., Sekar A., Williams E.. Techno-Economic Analysis of Supercritical Extraction of Rare Earth Elements from Coal Ash. Journal of Cleaner Production. 2018;189:539–551. doi: 10.1016/j.jclepro.2018.03.252. [DOI] [Google Scholar]
  48. El Batouti M., Al-Harby N. F., Elewa M. M.. A Review on Promising Membrane Technology Approaches for Heavy Metal Removal from Water and Wastewater to Solve Water Crisis. Water. 2021;13(22):3241. doi: 10.3390/w13223241. [DOI] [Google Scholar]
  49. Qalyoubi L., Al-Othman A., Al-Asheh S.. Recent Progress and Challenges of Adsorptive Membranes for the Removal of Pollutants from Wastewater. Part II: Environmental Applications. Case Studies in Chemical and Environmental Engineering. 2021;3:100102. doi: 10.1016/j.cscee.2021.100102. [DOI] [Google Scholar]
  50. Hammas-Nasri I., Elgharbi S., Ferhi M., Horchani-Naifer K., Férid M.. Investigation of Phosphogypsum Valorization by the Integration of the Merseburg Method. New J. Chem. 2020;44(19):8010–8017. doi: 10.1039/D0NJ00387E. [DOI] [Google Scholar]
  51. Buchner G. A., Zimmermann A. W., Hohgräve A. E., Schomäcker R.. Techno-Economic Assessment Framework for the Chemical IndustryBased on Technology Readiness Levels. Ind. Eng. Chem. Res. 2018;57(25):8502–8517. doi: 10.1021/acs.iecr.8b01248. [DOI] [Google Scholar]
  52. Lee J. C. K., Wen Z.. Pathways for Greening the Supply of Rare Earth Elements in China. Nat. Sustain. 2018;1(10):598–605. doi: 10.1038/s41893-018-0154-5. [DOI] [Google Scholar]
  53. Bailey G., Joyce P. J., Schrijvers D., Schulze R., Sylvestre A. M., Sprecher B., Vahidi E., Dewulf W., Van Acker K.. Review and New Life Cycle Assessment for Rare Earth Production from Bastnäsite, Ion Adsorption Clays and Lateritic Monazite. Resources, Conservation and Recycling. 2020;155:104675. doi: 10.1016/j.resconrec.2019.104675. [DOI] [Google Scholar]
  54. Bakas I., Hauschild M. Z., Astrup T. F., Rosenbaum R. K.. Preparing the Ground for an Operational Handling of Long-Term Emissions in LCA. Int. J. Life Cycle Assess. 2015;20(10):1444–1455. doi: 10.1007/s11367-015-0941-4. [DOI] [Google Scholar]
  55. Frischknecht, R. ; Jungbluth, N. ; Althaus, H.-J. ; Bauer, C. ; Doka, G. ; Dones, R. ; Frischnecht, R. ; Hischier, R. ; Hellweg, S. ; Humbert, S. ; Jungbluth, N. ; Köllner, T. ; Loerincik, Y. ; Margni, M. ; Nemecek, T. . Implementation of Life Cycle Impact Assessment Methods. Ecoinvent Report No. 3, v2.0; Swiss Centre for Life Cycle Inventories: Dübendorf, 2007.;https://esu-services.ch/fileadmin/download/publicLCI/03_LCIA-Implementation.pdf (accessed Feb 13, 2025). [Google Scholar]
  56. Paulillo A., Clift R., Dodds J., Milliken A., Palethorpe S. J., Lettieri P.. Radiological Impact Assessment Approaches for Life Cycle Assessment: A Review and Possible Ways Forward. Environ. Rev. 2018;26(3):239–254. doi: 10.1139/er-2018-0004. [DOI] [Google Scholar]
  57. Boulay A.-M., Bare J., Benini L., Berger M., Lathuillière M. J., Manzardo A., Margni M., Motoshita M., Núñez M., Pastor A. V., Ridoutt B., Oki T., Worbe S., Pfister S.. The WULCA Consensus Characterization Model for Water Scarcity Footprints: Assessing Impacts of Water Consumption Based on Available Water Remaining (AWARE) Int. J. Life Cycle Assess. 2018;23(2):368–378. doi: 10.1007/s11367-017-1333-8. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

es5c04952_si_001.pdf (928.8KB, pdf)

Data Availability Statement

The code used in the analysis is available at 10.5281/zenodo.15126344.

Articles from Environmental Science & Technology are provided here courtesy of American Chemical Society

RESOURCES