CHEMICAL ENGINEERING TRANSACTIONS

VOL. 81, 2020

A publication of

The Italian Association
of Chemical Engineering
Online at www.cetjournal.it

Guest Editors: Petar S. Varbanov, Qiuwang Wang, Min Zeng, Panos Seferlis, Ting Ma, Jiří J. Klemeš

ISBN 978-88-95608-79-2; ISSN 2283-9216

Comparative Life Cycle Assessment of three Recycling

Approaches for Electric Vehicle Lithium-ion Battery after

Cascaded Use

Yanqiu Tao*, Fengqi You

Smith School of Chemical and Biomolecular Engineering, Cornell University, Ithaca, New York, 14853, USA

yt554@cornell.edu

With increasing concerns on environmental impacts of retired lithium-ion batteries (LIBs) and supply risk of

critical materials, second life and recycling are considered as promising strategies to mitigate the environmental

impacts of retired automotive LIBs. In this life cycle assessment (LCA) study, we investigate environmental

benefits of second life and recycling methods on three types of widely adopted, commercialized automotive

LIBs, including lithium manganese oxide (LMO), lithium nickel manganese cobalt oxide (NCM) and lithium nickel

cobalt aluminum oxide (NCA). The results show that when subjected to second life, LIBs with both superior

energy storage capacity and energy density generate more environmental benefits than LIBs that is less

competitive in either working performance (31 % - 44 %, 16 % - 21 % and 50 % - 56 % reduction in environmental

impacts). This is in line with the current trend towards the development of electric vehicles with larger energy

storage capacity and longer battery life. We also found that direct cathode recycling outperforms

hydrometallurgical and pyrometallurgical recycling regardless of battery chemistry. Among all recycling

methods, only pyrometallurgical recycling is carbon intensive. The recovery of cathode active material and

metals, including copper and aluminum, are critical to the environmental performance of batteries.

1. Introduction

Owing to the raised concern in global warming, technology innovation, reduced cost and increased reliability in

safety and lifetime, the electric vehicle (EV) market grows exponentially starting from the year of 2010 and was

estimated to reach an annual supply of 40 GWh (Ciez and Whitacre, 2011). Given the commonly 8-year warranty

of EV, a huge amount of retired LIBs can be expected to be disposed of (Luo & Zeng, 2017). Previous studies

showed that lithium-ion batteries lost only around 20 % in their initial energy capacity at the end of their life in

EV (Wang et al., 2019), because of the rigid performance standards of EV (Neubauer and Pesaran, 2011).

However, less-demanding applications such as stationary energy storage systems (ESS) could serve as an

alternative to direct disposal (Zackrisson et al., 2010). Tesla has developed a Megapack made up of LIBs in

Australia and successfully facilitate in mediating the local electric grid and resulting in an annual saving of nearly

$40 MM (Tesla, 2019). Spent LIBs have the potential to be re-used in ESS to make profits, (Heymans et al.,

2014) and alleviate burdens metal supply (MaKinsey, 2018).

Previous work also proved environmental feasibility of cascaded use to offset energy consumption and carbon

footprint (Richa et al., 2017). However, the effect of cascaded use on different types of LIBs has not yet been

explored. Besides, the impact of cascaded use on different recycling methods has not been studied. Other than

global warming potential (GWP) and cumulative energy demand (CED), a more comprehensive environmental

profile such as ReCiPe and TRACI is under investigation. In this study, we conduct a comparative LCA to assess

the GWP, CED and ReCiPe endpoint indicators for LMO, NCM and NCA LIBs with first use in EV and second

use in ESS before subsequently put to one of the three recycling methods, namely hydrometallurgical,

pyrometallurgical and direct cathode recycling.

DOI: 10.3303/CET2081188

Paper Received: 25/03/2020; Revised: 25/06/2020; Accepted: 26/06/2020
Please cite this article as: Tao Y., You F., 2020, Comparative Life Cycle Assessment of three Recycling Approaches for Electric Vehicle Lithium-
ion Battery after Cascaded Use, Chemical Engineering Transactions, 81, 1123-1128 DOI:10.3303/CET2081188

1123

2. Material and methods

In this study, we conducted a “cradle-to-cradle” LCA to estimate the GWP and CED associated with the

production, consumption and end-of-life of three automotive lithium-ion batteries, namely LMO, NCM622 and

NCA, after cascaded use in the stationary energy storage system. In addition, to obtain a more comprehensive

environmental profile, 3 ReCiPe endpoint indicators from a hierarchal perspective were evaluated to assess

damage to ecosystems, damage to human health and damage to resource availability. All values of GWP, CED

and ReCiPe scores were sourced from Ecoinvent V3.6 (Wenet et al., 2016).

2.1 Life cycle inventory

The environmental impact associated with cathode active material of LMO was obtainable directly through the

Ecoinvent database while the environmental impact associated with avoided Ni(OH)2 production and cathode

active material of NCM and NCA production were not available. Instead, the upstream materials and energy for

producing Ni(OH)2 sourced from (Majeau-Bettez et al., 2011) and the NCM and NCA cathode active materials

were acquired from the GREET model (Dunn et al., 2012). The production of NCM and NCA cathode active

material requires material inputs of lithium source, transit metal sources, deionized water, NH4OH and NaOH,

energy inputs and waste sludge treatment (Zhao et al., 2019). It is worth to mention that the product yields are

all assumed to be 100 % since the production is modelled based on industrial data which is designed to minimize

material loss and result in nearly 100 % product yields. In addition, the by-product of sodium sulfate was

excluded from the system boundary using the concept of “recycled content” approach.

Working parameters, such as energy consumption rate in electric vehicle and energy storage capacity, are

extracted from the BatPac model. Other working parameters, such as initial battery discharging efficiency,

battery discharging efficiency after electric vehicle use and after stationary ESS use, the lifetime of the EV

battery pack and stationary ESS battery pack, roundtrip and transmission efficiency for stationary ESS, and

Depth-of-discharge (DoD) of stationary ESS battery pack, are extracted from literature (Richa et al., 2017).

Automotive use is assumed to be 100,000 miles over an 8-year timeframe for both use scenarios of all types of

batteries, and will the discharging efficiency would decrease from 100 % to 80 % (Väyrynen and Salminen,

2012). Starting from an initial discharging efficiency of 80 %, Stationary ESS use is assumed to operate one

charge/discharge cycle on a daily basis for 10 years with an end-of-life (EOL) discharging efficiency of 65 %.

After retired from electric vehicles, battery cells will be tested to validate their residual energy capacity and

consumed electricity. (Richa et al., 2017) Life-extending battery cells are assumed to be repurposed into new

battery packs with an energy storage capacity of 450 kWh, aiming for their second use in stationary ESS. All

battery module components and part of battery pack components such as module compression plates and steep

straps are reused, while coolant, battery management system (BMS), module interconnects and a tri-layer

battery jacket are replaced. The replaced components will be sent to the EOL phase. Notably, BMS is assumed

to be 1.5 % of the total mass of one repurposed battery pack, based on the fact that BMS of battery packs

designed for stationary ESS is not as demanding as those for the electric vehicles, and less advanced printed

wiring board design and less complicated wiring would be applied. A cabinet made of mild steel sheet was used

as the pack casing instead of the tri-layer jacket. The life cycle inventory for two use phase scenarios and three

EOL scenarios will be detailed in the following two sections.

2.2 Use phase scenarios

Two use phase scenarios are assessed: one is electric vehicle use only, and the other is second use in

stationary energy storage system after EV use. We adopt a combination of two functional units, namely

functional unit LCED (1 kWh of life-cycle electricity delivered) and functional unit ESC (1 kWh of battery energy

storage capacity). Both functional units measure the function of lithium-ion batteries, although from different

perspectives. In terms of comparing life-cycle environmental impacts of batteries with EV use scenario and

cascaded use scenario, the functional unit LCED delivered focus on evaluating the batteries’ life-cycle

performance of delivering electricity. However, in terms of comparing across different types of batteries, the

functional unit LCED is considered insufficient due to its incapability to reflect the energy storage capacity of

different batteries. The functional unit of ESC, which is widely adopted in previous LCA studies of lithium-ion

batteries, is added in addition to the functional unit of LCED to complement the shaded view.

2.3 End-of-life scenarios

EOL of lithium-ion batteries involves dismantling, material production, energy generation, incineration,

combustion, waste sludge treatment, energy and material recovery. The environmental impact associated with

energy and material recovery is considered as avoided burden and is reported as a reduction in carbon footprint

and primary energy consumption. A comprehensive material flow analysis is performed and coupled with LCA

(Gao et al., 2018). Three EOL scenarios, namely hydrometallurgical, pyrometallurgical and direct cathode

1124

recycling, are investigated. Notably, for all EOL scenarios, the spent LIB pack was first discharged and

dismantled. In this step, the plastics were landfilled. The metals such as aluminum, chromium steel and copper

are recycled with a loss rate of 5 %, 1.3 % and 8 %. Waste steel is landfilled and waste fiber glasses are

incinerated. Printed wiring boards are prepared for further metallurgical treatment. (Wenet et al., 2016) Coolant

is incinerated as hazardous waste. After disassembly, cathode, anode and electrolyte are left. Besides, plastics,

aluminum and copper, were left for the scenario of pyrometallurgical recycling. Hydrometallurgical recycling

aims at recovering metals using aqueous chemistry, which involves steps of leaching, solvent extraction and

precipitation. Under this scenario, the leftovers are soaked in N-Methyl-2-Pyrrolidone (NMP) and crushed to

dissolve the binder Polyvinylidene fluoride (PVDF) and separate the active materials from the current collectors.

(Dunn et al., 2012) Later, solids such as graphite, carbon black, copper and aluminum are filtered out, the carbon

black and graphite are completely combusted to CO2, and the filtrate is calcined and grinded. Leaching and

precipitation processes are used to extract the valuable metals and recover the cathode active material. The

leaching efficiency is 95 % of Li for LMO, 99.7 % of Li, Ni, Co, Mn for NCM, and 80 % of Li, 99.8 % of Ni, 95.6

% of Co and 99.5 % of Al for NCA.

Direct cathode recycling process focuses on recover the cathode active material and electrolyte with limited

processing. After dismantling, cathode, anode and electrolyte are crushed and electrolyte is extracted by liquid

CO2. Experiments showed that the recovery rate of LiPF6, ethylene carbonate and dimethyl carbonate are 100

%, 75.76 % and 81.71 %. (Nowak and Winter, 2017) After the electrolyte was extracted and recovered, the

residues underwent steps of size reduction and final separation which both requires electricity. Then, they

undergo NMP soaking, hydrothermal treatment and annealing process to recover the cathode active materials.

Pyrometallurgical recycling is developed by Umicore and is designed to recover Nickel and Cobalt by forming

an alloy through a three-stage smelting. (Cheret and Santen, 2007) After disassembly, the left spent LIB

modules is fed directly into the smelting furnace together with slag-forming materials. First, the electrolyte is

evaporated and the plastics are melted. Subsequently, blast furnace slag and alloy are formed. The gaseous

phase is post-combusted in a combined heat and power generation (CHP) facility and complete combustion

was assumed. The alloy is further leached and precipitated to recover copper, goethite, Ni(OH)2 and cobalt ion

solution.

3. Results and discussion

The results of carbon footprint, primary energy consumption and ReCiPe endpoint indicators are presented

based on distinguished use scenarios and EOL scenarios.

3.1 Environmental benefits of introducing second life to batteries

To assess the implications of second life application to LIBs from the environmental perspectives, we compare

the life cycle environmental impacts between the two use scenarios (with and without second life) across LMO,

NCM and NCA LIBs using functional unit LCED and ESC as shown in Figure 1a and 1b. Life cycle

environmental impacts are quantified in the form of ReCiPe scores. Environmental scores are calculated

for each ReCiPe midpoint indicators by translating complicated life cycle inventories (LCIs) into 18 impact

categories. Later, ReCiPe scores of 3 endpoint indicators are summarized into the ReCiPe total scores and

are shown as data labels on top of each bar in Figure 1The EOL recycling method is kept the same when

this comparison is made (Yang et al., 2017). From the perspective of life-cycle devotion to electricity

delivery (Figure 1a), NCM LIBs perform the worst without second life (16 %, 76 % and 83 % more than the

total ReCiPe score of LMO LIBs, and 27 %, 29 % and 40 % more than that of NCA LIBs for

hydrometallurgical, direct cathode and pyrometallurgical recycling). In the presence of cascaded use, the

environmental performance of NCM LIBs become drastically favorable and comparable with those of LMO and

NCALIBs.

This notable reduction in the ReCiPe total scores of NCM LIBs can be mainly attributed to the highest amount

of electricity delivered during second life. Remarkably, this can be fundamentally attributed to three factors: (1)

the superior energy storage capacity of the NCM LIBs (88.17 kWh per EV battery pack) over the LMO and NCA

LIBs (23.53 and 52.94 kWh per EV battery pack), which results in a comparable level of electricity delivery

during automotive use but a significantly higher amount of electricity delivery during second life; (2) the

outstanding battery pack energy density of NCM (181 Wh/kg) that conveys higher energy storage capacity to

the same weight of environmentally-expensive LIB pack, compared to the LMO (159 Wh/kg) and NCA (172

Wh/kg) LIBs; (3) the overwhelming environmental impacts associated with a large amount of electricity delivery

during the use phase (accounts for 55 % - 62 % and 24 % - 31 % of the total ReCiPe score for NCM LIBs

with and without second life). The third factor is also responsible for the increase in ratios of scores for

“damage to human health” and “damage to resource availability” and decrease in that for “damage to

ecosystems”. Since the electricity grid in New York State (NYS) is mainly powered by natural gas,

nuclear and hydroelectric,“damage to human health”, “damage to resource availability” and “damage to
ecosystems” account for 29 %, 29 % and 42 % of the total ReCiPe score associated with the use phase.

1125

When environmental impacts associated with use phase increases, the shares of ReCiPe scores for three

endpoint indicators get closer to that of the use phase.

Figure 1: Score of ReCiPe endpoint indicators comparing battery chemistries, use scenarios and EOL scenarios

based on a. functional unit LCED and b. functional unit ESC.

Switching to the perspective of energy storage capacity, NCM LIBs achieve outstanding environmental

performance while LMO LIBs achieve the worst performance regardless of the use scenarios. This is because

of the superiority in both battery pack energy storage capacity and energy density. Notably, the difference

between use scenarios is not significant, because functional unit ESC does not address any function related to

the second life.

3.2 Environmental impacts of battery recycling methods on second life

The environmental impacts of battery recycling methods on second life are shown in Figure 1. The EOL phase

appears to be the leading contributor for the total ReCiPe score. The direct cathode recycling outperforms the

other two recycling methods regardless of battery chemistry, while hydrometallurgical recycling outperforms

pyrometallurgical recycling only for NCM and NCA LIBs. Due to the utility of environmentally expensive leachate

citric acid and the manganese source Mn2O3, hydrometallurgical recycling of LMO underperforms

pyrometallurgical recycling. Instead, EOL phase with positive ReCiPe score is much less influential than EOL

phase with negative ReCiPe score. Hydrometallurgical and direct cathode recycling result in comparable

negative ReCiPe scores, while hydrometallurgical recycling leads to a higher positive ReCiPe score compared

to the direct cathode recycling. On the contrary, pyrometallurgical recycling leads to both the most positive

ReCiPe score and the least negative ReCiPe score for all LIBs.

Both the inter-battery chemistry and the inter-EOL scenario variances in the total ReCiPe scores become less.

In particular, for use scenario with and without second life, the inter-EOL scenario standard deviation change

from 0.013, 0.010, 0.005 to 0.006, 0.002, 0.002 for LMO, NCM and NCA LIBs. The inter-battery chemistry

standard deviation changed from 0.008 – 0.021 to 0.001 – 0.005.

3.3 Life-cycle carbon footprint and primary energy consumption

The GWP and CED are two important metrics to evaluate the climate change mitigation potential and energy

performance of introducing second life and recycling into batteries’ life cycle. The results of GWP and CED are

compared from the perspective of life cycle delivered electricity and energy storage capacity as demonstrated

in Figure 2a and 2b using stacked bars and red rhombus.

Figure 2a and 2b show that from the perspective of functional unit LCED, the GWP and CED result in

environmental benefits of introducing the second life application to LIBs is consistent with that of ReCiPe results:

NCM LIBs without second life are the most carbon intensive among all types of LIBs (the GWP of NCM LIBs is

twice more than that of LMO LIBs), while under the cascaded use scenario, carbon footprint of NCM LIBs

become comparable to that of NCA LIBs and only 32 % - 51 % more than that of LMO LIBs. In particular, the

difference between life-cycle carbon footprint of NCM and LMO LIBs drops from 0.092 – 0.188 kg CO2 eq. to

0.009 – 0.023 kg CO2 eq. after adopting second life. Compared with use scenario of only automotive use, LMO,

NCM and NCA LIBs with second life achieve 9 % - 15 %, 31 % - 44 % and 19 % - 30 % reduction in life cycle

carbon footprint. The reductions in primary energy consumption are less distinct in that LMO, NCM and NCA

reduce 3.6 % - 4.8 %, 15.6 % - 21.3 % and 8.9 % - 13.2 % of CED when adopting second life. This suggests

1126

that while LMO remains to be the least carbon and energy intensity, the adoption of second life can pull down

the high carbon and energy intensity of LIBs such as NCM and NCA substantially to an acceptable level. In

other words, the adoption of second life can relieve many of the concerns towards the commercialization of lab-

scale batteries with superior working performance but costly environmental impacts.

Compared to the ReCiPe results, cascaded use accounts for a higher share in the total life cycle GWP (81.4 %

- 87.1 %, 74.1 % - 83.6 %, and 74.0 % - 83.3 % for LMO, NCM and NCA LIBs) and CED (93.6 % - 94.9 %, 90.4

% - 92.8 % and 90.2 % - 92.8 % for LMO, NCM and NCA LIBs). Notably, the shares of automotive use in the

GWP and CED values associated with cascaded use for LMO, NCM and NCA LIBs is the same as that of the

ReCiPe result (46.6 %, 22.5 % and 32.6 %). This revalidates the conclusion that batteries with larger energy

density could be more environmentally sustainable if subjected to second life. In consequence of the large

proportion of CED related to the use phase, other phases appear not as significant as what is shown in GWP

results.

Instead of being the most influential contributor to the total ReCiPe score, the EOL phase appears to be the

second leading contributor for both carbon footprint and primary energy consumption. Among the three EOL

scenarios, direct cathode recycling appears to be the least carbon intensive for all LIBs. On the contrary,

hydrometallurgical recycling of LMO LIBs surpasses direct cathode recycling in terms of primary energy

consumption. Notably, only pyrometallurgical recycling results in a net positive carbon footprint, while it appears

to be slightly energy saving. This result is different from that of the ReCiPe score, where all three recycling

methods can lead to net negative ReCiPe scores that have much larger impacts on avoiding batteries’

environmental burdens. The result suggests that instead of using metrics such as carbon footprint and primary

energy consumption, the environmental benefits of introducing second life and recycling into batteries’ life cycle

can be better revealed by the systematic life cycle environmental impact assessment.

By further disaggregating the EOL phase into positive and negative portions, it can be observed that direct

cathode recycling results in the least positive GWP, while at the same time avoiding the most GWP for all three

types of LIBs. Nevertheless, pyrometallurgical recycling leads to both the most positive GWP and the least

negative GWP. Similarly, direct cathode recycling and pyrometallurgical recycling lead to the most and the least

negative CED. The positive CED results are more complicated: first, hydrometallurgical recycling has the most

positive CED for NCM and NCA LIBs, but LMO LIBs has the least positive CED by using hydrometallurgical

recycling; second, direct cathode recycling leads to the most positive CED for LMO LIBs and the least positive

CED for NCA LIBs; the pyrometallurgical recycling leads to the least positive CED for NCM LIBs.

Figure 2: Breakdown of a. carbon footprint and b. primary energy consumption for LMO, NCM622 and NCA

LIBs with three EOL scenarios and two use scenarios

As a consequence of the large proportion of GWP and CED related to the use phase, other phases appear not

so significant as shown in ReCiPe results. The material production and assembly for cell and pack accessories

are ranked third and fourth in their contributions to carbon footprint and primary energy consumption. Among

the three types of LIBs, NCM cell production appears to be the most carbon and energy intensive (1 % - 5 %

and 108 % - 118 % higher than GWP and CED of NCA and LMO cell production). The pack accessories

production and assembly do not show a significant difference in GWP between these LIBs. Module accessories

production and assembly, repurpose and transportation account for no more than 5 %, 3.5 % and 1 % of the life

cycle GWP and CED. They become relatively negligible.

From the perspective of functional unit ESC, NCM LIBs outperforms LMO and NCA LIBs in terms of life-cycle

carbon footprint and primary energy consumption regardless of EOL scenarios. Direct cathode recycling

1127

remains to have the smallest carbon footprint and performs slightly worse than hydrometallurgical recycling in

terms of primary energy consumption.

4. Conclusions

The comparative LCA illustrated the advantages of second use for spent LIBs from automotive use, especially

on those with superior performance such as high energy storage capacity and high energy density, i.e. NCM

LIBs in this study. Concerns towards environmental impacts can be relieved in that the adoption of cascaded

use reduces 50 % - 56 % of ReCiPe score, 31 % - 44 % of carbon footprint and 16 % - 21 % of primary energy

consumption for NCM LIBs and pulled down the environmental impacts of NCM LIBs to a level that is

comparable to that of other batteries with less environmental impacts. Comparison of recycling methods points

out the necessity to recover cathode active material that is shown to be the most environmental expensive

process throughout the batteries’ life cycle. Direct cathode recycling outperforms hydrometallurgical and

pyrometallurgical recycling in all impact categories, however, fell short of expected advantages due to the

overwhelming energy consumption of liquid CO2 production, suggesting that efficiency improvement and the

alternatives to the current environmental expensive process or materials could be further studied in the future

battery recycling research.

References

Azevedo M, Campagnol N., Hagenbruch T., Hoffman K., Lala A, Ramsbottom O., 2018. Metals and mining,

lithium and cobalt – a tale of two commodities. McKinsey&Company.

Cheret D., Santen S., 2007, Battery recycling. US patent 7,169,206 B2.

Ciez R.E., Whiteacre J.F, 2019, Examining different recycling processes for lithium-ion batteries, Nature

Sustainability, 2, 148-156.

Dunn J.B., Gaines L., Barnes M., Sullivan J., Wang M.Q., 2012. Material and energy flows in materials

production, assembly, and end-of-life stages of the life cycle of lithium-ion batteries. ANL/ESD/12-3, Argonne

National Laboratory. <https://greet.es.anl.gov/publicationlib-lca> accessed 21.03.2020

Gao J., You F., 2018. Dynamic material flow analysis-based life cycle optimization framework. ACS Sustainable

Chemistry & Engineering, 6, 11734-11752.

He C., You F., 2016. Deciphering the true life cycle environmental impacts and costs of the mega-scale shale

gas-to-olefins projects in the United States. Energy & Environmental Science, 9, 820-840.

Heymans C., Walker S.B., Young S.B., Fowler M., 2014. Economic analysis of second use electric vehicle

batteries for residential energy storage and load-levelling. Energy Policy, 71, 22–30.

Luo J., Zeng Y., 2017. Research on development of the lithium-Ion battery. Chemical Engineering Transactions,

62, 97-102.

Majeau-Bettez G., Hawkins T.R., Strømman A.H., 2011. Life cycle environmental assessment of lithium-ion and

nickel metal hydride batteries for plug-in hybrid and battery electric vehicles. Environmental Science &

Technology, 45(10), 4548-4554.

Neubauer J, Pesaran A, 2011. The ability of battery second use strategies to impact plug-in electric vehicle

prices and serve utility energy storage applications. J Power Sources, 196(23),10351–10358.

Nowak S., Winter M., 2017. The role of sub- and supercritical CO2 as “processing solvent“ for the recycling and

sample preparation of lithium-ion battery electrolytes. Molecules, 22, 403.

Richa K., Babbitt C.W., Nenadic N.G., Gaustad G., 2017. Environmental trade-offs across cascading lithium-ion

battery life cycles. The International Journal of Life Cycle Assessment, 22, 66-81.

Tesla, Introducing megapack: utility-scale energy storage, 2019. <tesla.com/blog/introducing-megapack-utility-

scale-energy-storage> accessed 26.3.2020.

Väyrynen A., Salminen J., 2012. Lithium-ion battery production. The Journal of Chemical Thermodynamics, 46,

80-85.

Wang T., Kang L., Liu Y., 2019. Multi-objective optimisation method for identifying retired points of electric

vehicle batteries. Chemical Engineering Transactions, 76, 925-930.

Yang M., You F., 2017. Comparative techno-economic and environmental analysis of ethylene and propylene

manufacturing from wet shale gas and naphtha. Industrial & Engineering Chemistry Research, 56, 4038-

4051.

Yue D., Pandya S., You F., 2016. Integrating Hybrid Life Cycle Assessment with multiobjective optimization: a

modeling framework. Environmental Science & Technology, 50, 1501-1509.

Zackrisson M., Avellán L., Orlenius J., 2010. Life cycle assessment of lithium-ion batteries for plug-in hybrid

electric vehicles – critical issues. Journal of Cleaner Production, 18, 1519-1529.

Zhao S., You F., 2019. Comparative life-cycle assessment of Li-ion batteries through process-based and

integrated hybrid approaches. ACS Sustainable Chemistry & Engineering, 7, 5082-5094.

1128