Available online at www.sciencedirect.

com Current Opinion in

ScienceDirect Green and Sustainable Chemistry

Recent progress and future prospects of silicon solar

module recycling
Meng Tao, Theresa Chen, Natalie Click and Randall Adcock

Abstract This includes mainly the aluminum (Al) frame and

Silicon solar modules are only 10–15 wt% circular with today’s sometimes the copper (Cu) wires in the junction box.
recycling technologies. A 90 wt% circularity requires that all the The other 85e90 wt% of the materials are either land-
inorganic materials in silicon modules be recovered for reuse in filled or sometimes downcycled to a construction ma-
solar or similar applications. Major technical barriers to a 90 wt terial containing glass, silicon solar cells, and polymers.
% circularity for silicon modules include: 1) removal of the
fluoropolymer back sheet; 2) detachment of silicon cells from It has been envisioned that solar energy will become a
glass; 3) removal of the encapsulant on silicon cells; and 4) main energy source for the world by 2050. Such a goal
mild chemistry and minimization of chemical waste along with requires 30e100 TW peak (TWp) of solar modules to be
high material recovery rates. Noticeable progress in recycling produced and installed in the next 28 years [2,3]. By the
technologies in the last few years includes: 1) mechanical end of 2021, the global cumulative installation of solar
milling to remove the fluoropolymer back sheet; 2) laser modules was 0.94 TWp [4], so the installation needs to
debonding of the encapsulant from silicon cells; 3) dissolution expand about 50 times by 2050. Silicon solar modules have
of the encapsulant with a base; 4) mild chemistry for silver and a typical, manufacturer-guaranteed lifetime of 25 years. A
lead recovery; and 5) regenerative chemistry to reuse some of steady-state solar installation of 50 TWp means that 2
the chemicals in silicon cell recycling. TWp of solar modules would reach their end-of-life each
year. A typical 300 W peak (Wp) silicon module weighs
Addresses
about 18 kg, so the 2 TWp/year end-of-life silicon modules
Arizona State University, Tempe, AZ 85281, USA mean over 100 megatonnes (Mt) of solar waste a year. The
100 Mt/year solar waste does not include balance-of-
Corresponding author: Tao, M. ([email protected]) system waste such as the mounting components, elec-
trical wiring, power converters, and storage batteries. With
Current Opinion in Green and Sustainable Chemistry 2023, everything included, the solar waste could exceed 200 Mt/
44:100863
year shortly after 2050. To compare, the total municipal
This review comes from a themed issue on Recycling of solar panels solid waste in the US was 265 Mt in 2018 [5].
and windmills (2023)
Edited by Peter Majewski and Pablo Dias This paper analyzes recent progress and future pros-
Available online 23 August 2023 pects to achieve a 90e95 wt% circularity for silicon
For complete overview of the section, please refer the article solar modules. This means that 90e95 wt% of the
collection - Recycling of solar panels and windmills (2023) materials in silicon modules are recovered for reuse in
https://doi.org/10.1016/j.cogsc.2023.100863 solar or similar applications, not downcycled for less-
2452-2236/© 2023 Elsevier B.V. All rights reserved. demanding applications. It is recognized that some
materials in silicon modules are difficult to recycle,
especially the polymers including the encapsulant and
Keywords
Silicon solar module recycling, Chemical recycling, Lead, Silver, Silicon,
back sheet. This analysis will focus on recent work
Encapsulant removal. since 2018 in both the research community and private
sector. For reviews on silicon module recycling tech-
nologies before 2018, the readers are referred to
Introduction
Refs. [6e9].
Solar modules have been branded as a “green” technol-
ogy. They are green in the sense of utilizing a sustainable
energy source, i.e., solar energy, to generate electricity, Structures of silicon solar modules and
but silicon (Si) solar modules are not green in the sense cells
of a circular economy. Today most end-of-life solar Figure 1(a) shows schematically the cross section of the
modules are sent to landfills. First Solar provides recy- most common commercial silicon solar module today.
cling services for their cadmium telluride modules, but The major components in silicon modules include the
silicon modules have a 95% market share as of 2021 [1]. front glass sheet, aluminum frame, silicon solar cells,
Even for those modules which go through today’s recy- junction box on the back (not shown in Figure 1(a)), and
cling processes, they are only 10e15% circular by weight polymers including the encapsulant, sheath for copper
(wt%) meaning that only 10e15 wt% of the materials wires, casing for the junction box, and back sheet. The
contained in silicon modules are recovered for reuse. most common encapsulant is ethylene vinyl acetate

www.sciencedirect.com Current Opinion in Green and Sustainable Chemistry 2023, 44:100863

2 Recycling of solar panels and windmills (2023)

Figure 1

Schematic cross sections of (a) silicon solar module without the junction box; (b) aluminum back-surface field cell; and (c) passivated emitter rear contact cell.

(EVA) but polyolefin may replace EVA in the future.  Tin

The back sheet is typically polyvinyl fluoride, a fluo-  Copper ribbons
ropolymer. The glass has a low iron content (below  Silicon
100 parts per million (ppm)) for a high transmittance.
The junction box contains copper wires. Therefore, the
materials worth recovering at the module level are: Out of the five metals in silicon solar cells, two are in
the must-recover category: silver and lead [7]. Silver is
 Aluminum frame a main financial incentive for recyclers. Each
 Low-iron glass 156  156 mm2 Al-BSF cell contains about 0.1 g of
 Copper wires silver. At today’s silver price and with 100% recovery,
 Silicon cells the silver generates roughly 8¢/cell. Lead is toxic, and
lead in the presence of EVA is a particularly bad
The hydrocarbon polymers can be in principle recovered combination. When landfilled, EVA will decompose to
as a fuel. The only material which must be landfilled is acetic acid under moisture, ultraviolet radiation, and
the fluoropolymer back sheet since there is currently no heat. Our group demonstrated that acetic acid is an
better method to deal with it. effective leaching agent for lead recovery from silicon
cells [10]. A 300 Wp Al-BSF module contains about
Figure 1(b) is the schematic cross section of the 15 g of lead.
aluminum back-surface field (Al-BSF) cell which
dominated the solar market till about 2020, so this is the In early 2022, the US Environmental Protection Agency
silicon solar cell recyclers will process for the next 20 revised its guidelines on end-of-life management of solar
years. Figure 1(c) is the passivated emitter rear contact modules [11]. Now they recommend the toxicity char-
(PERC) cell which dominates the current solar market, acteristic leaching procedure (TCLP) to determine if
so recyclers will see a small but increasing portion of this end-of-life solar modules are hazardous waste. One
cell. The two cells have a similar structure, except that concern about the TCLP test is that it may not be
the PERC cell has a stack of aluminum oxide and silicon capable of revealing the true hazard of solar modules.
nitride (Al2O3/Si3N4) layers sandwiched between the The solder in silicon modules is sealed under the
silicon wafer and aluminum back electrode. The two encapsulant. If the encapsulant is intact, the TCLP test
cells can be recycled with the same process. may detect little lead. However, no encapsulant will last
forever when landfilled. When it decomposes, the solder
In Figure 1(b) and (c), the front electrode and soldering is exposed. This is the moment to see the true hazard of
pads on the back are made of silver (Ag). Before module silicon modules.
production, multiple silicon cells are soldered in series with
copper ribbons for electrical connection. The solder is an Each 300 Wp Al-BSF module contains about 0.55 kg of
alloy of about 65% tin (Sn) and the rest lead (Pb). The silicon. Three different qualities of silicon can be recov-
aluminum oxide and silicon nitride layers are extremely ered [7]:
thin and negligible by weight, with no value. The
aluminum back electrode is difficult to recover with little  Ferrosilicon with a minimum silicon content of 75%
value. Therefore, five metals are worth recovering from  Metallurgical-grade silicon with a minimum silicon
silicon cells. content of 98%
 Solar-grade silicon with a minimum purity of
 Silver 99.9999% or 6N
 Lead

Current Opinion in Green and Sustainable Chemistry 2023, 44:100863 www.sciencedirect.com

 Progress in silicon solar module recycling Tao et al. 3

The silicon cells, once detached from the glass with all copper wires, aluminum frame, low-iron glass, and
the metals, dielectrics, and some EVA on them silicon cells. The recovered copper, aluminum, and
(Figure 1(b) and (c)), can be sold directly as ferrosilicon as glass, which account for about 90% of the module
they contain about 80% silicon. Therefore, the recovery weight, can be remelted and reused in new silicon
rate of ferrosilicon can reach 120% of the silicon contained modules. The silicon cells (about 5% of the module
in a module [12]. To recover metallurgical-grade silicon, weight) are sent to chemical recyclers for further
the EVA, metals, and dielectrics must be removed so the processing. At the end of this stage, some polymers
maximum recovery rate is limited to 100%. To extract (3e6 wt% depending on the amount of hydrocarbon
solar-grade silicon, the heavily doped emitter (the orange polymers recovered as a fuel) are landfilled.
section in Figure 1(b) and (c)) and back-surface field (the  The second stage deals with chemical recycling of the
dark green section in Figure 1(b) and (c)) must be silicon cells to recover the metals including silver,
removed in addition to the EVA, metals, and dielectrics. lead, silicon (either metallurgical grade or solar grade),
Only the p-type base (the light green section in and maybe tin and copper ribbons. Technically it is
Figure 1(b) and (c)) qualifies as solar-grade silicon, so the possible to recover tin and copper ribbons, but re-
recovery rate of solar-grade silicon is limited to about 90%. cyclers may opt not to for financial reasons, as recovery
of a metal incurs cost. This stage requires some nasty
Recovery of solar-grade silicon is preferred as it adds, at chemicals and generates chemical waste.
today’s price for solar-grade silicon, 24¢/cell in revenue
for each 156  156 mm2 cell. For ferrosilicon and
metallurgical-grade silicon, the revenue is only 1e2¢/ Some of the major technical barriers to realizing the
cell. However, there are significant logistic and technical recycling process in Figure 2 include:
challenges to recover solar-grade silicon, as discussed
later in this paper. Therefore, it is possible that recyclers  Removal of back sheet: There is currently no com-
will stop with metallurgical-grade silicon [12]. mercial tool to remove the fluoropolymer back sheet,
which becomes an issue in subsequent processing. If a
Table 1 lists the bill of materials for a typical silicon solar thermal process is applied, e.g., pyrolysis of EVA, the
module. Glass and aluminum account for 85% of the back sheet releases a fluorine exhaust [13]. If a
module weight. The silicon solar cells including silver, chemical process follows, e.g., dissolution of EVA, the
lead, tin, aluminum back electrode, and copper ribbons back sheet remains undissolved and floats in the
are about 4% of the module weight, but they contain the chemical tank, interfering with the process.
two must-recover metals: precious silver and toxic lead.  Detachment of silicon cells from glass: Significant
Recovery of glass, aluminum, silicon, copper, tin, lead, efforts have been devoted to an effective detachment
and silver leads to a 90 wt% circularity. If some of the technology. It is reminded that any pieces of silicon
hydrocarbon polymers are recovered as a fuel, silicon cells in the recovered glass cullet render it unfit for
modules would become about 95 wt% circular. new glass products as silicon never melts during glass
production. It is also desirable to have silicon cells
Prospects of a proper recycling process for recovered glass-free, as glass interferes with the
silicon modules chemical cell recycling process.
Consensus seems to be emerging in the solar commu-  Removal of EVA: The EVA on silicon cells must be
nity that a proper recycling process for silicon modules removed to expose the cells for metal leaching. Py-
involves two stages [6,7,9] as shown in Figure 2: rolysis of EVA requires a large furnace to process
hundreds or even thousands of modules in each batch,
 The first stage involves physical disassembly of silicon but such a furnace is not commercially ready. More-
modules to recover the bulky materials including the over, pyrolysis emits carbon dioxide (CO2). An alter-
native is to chemically dissolve EVA, which is a
Table 1
difficult task for a crosslinked polymer.
 Chemical cell recycling: It is difficult to recover the
Bill of materials in a typical silicon solar module [8]. metals in silicon cells through physical methods, but
Material Weight % some of the chemicals used in metal recovery are nasty
[6e9]: nitric acid (HNO3), hydrofluoric acid (HF),
Glass 75
Al 10
and hydrochloric acid (HCl). More mild chemicals and
Hydrocarbon polymers 7 methods to minimize chemical waste are desirable,
Si cells 4 but they must achieve a high recovery rate for lead
Back sheet 3 (100%) and ideally for other metals as well.
Cu 1
Sn 0.15
Pb 0.1
Ag 0.03 These barriers should be the focus of research. In
particular, removal of EVA is currently the biggest

www.sciencedirect.com Current Opinion in Green and Sustainable Chemistry 2023, 44:100863

4 Recycling of solar panels and windmills (2023)

Figure 2

A proper recycling process for silicon solar modules: physical module disassembly first and then chemical cell recycling.

barrier as it is a prerequisite for chemical cell recycling. et al. [13] reported a mechanical milling method to
The next major barrier is to reduce the number, quan- remove the back sheet, which seems attractive.
tity, and hazard of the chemicals used for cell recycling,
while maintaining high material recovery rates. The There are currently two approaches to detach silicon
following sections discuss the two stages of silicon cells from glass, with the back sheet attached to the
module recycling, with a separate section dedicated to silicon cells in most cases. The more successful
EVA removal. approach is knife cutting through the EVA between the
glass and silicon cells. Many studies also take a crushing
and physical separation approach.
Stage 1: physical module disassembly
Today all the recyclers strip off the aluminum frame. Knife cutting of EVA
Some also cut off the junction box. There are two Knife cutting requires an elevated temperature to
methods to remove the plastics and extract the copper soften the EVA. NPC Inc. in Japan has a commercial tool
wires. The junction box can be sent to a copper smelter which employs a heated steel blade to cut through the
where it goes through several high-temperature pro- EVA [14]. A European Union funded project employs
cesses of about 1500  C and an electrorefining process to infrared lamps to heat the module before knife cutting
recover copper. The alternative is to mechanically of the EVA [15]. The NPC tool does a good job in
remove the plastic sheath and casing, and the recovered separating glass and silicon cells; it creates a clean glass
copper goes through one high-temperature process to sheet with some EVA and a sheet of silicon cells lami-
remelt for new copper wires. The plastic casing and nated together by EVA and back sheet. There is no
sheath, if they do not contain chlorine or fluorine, can be commercial tool based on the infrared lamp technology.
collected as a fuel.
There are still questions about these knife cutting tools:
After the removal of junction box and aluminum frame,
the remaining module has a layered structure  They do not remove the back sheet, which is still
(Figure 1(a)). The next step is to separate the glass, attached to silicon cells.
silicon cells, and ideally the back sheet as well. However,  The knife is made of steel so it may increase the iron
no recyclers today separate the back sheet from mod- content in the recovered glass cullet. Currently the
ules. In principle, the fluoropolymer back sheet can be recovered glass is not used for new solar glass sheets.
peeled off when the EVA is heated and softened, but To do that, the iron content in the recovered glass
there is no commercial tool to perform the job. Fiandra cullet must be below 100 ppm.

Current Opinion in Green and Sustainable Chemistry 2023, 44:100863 www.sciencedirect.com

 Progress in silicon solar module recycling Tao et al. 5

 The lifetime of the steel blade is limited as it endures Stage 1.5: removal of EVA
wear and tear against glass. Removal of EVA plays a dual role in silicon module
recycling:
Crushing and physical separation
With the junction box and aluminum frame gone, most  When applied to silicon modules, it enables detach-
of the recyclers today crush the remaining module which ment of silicon cells from glass. A subsequent step is
contains largely glass with silicon cells and polymers. required to separate silicon cells and glass.
Often the crushed modules are sent to landfills as glass  When applied to detached silicon cells, it exposes
manufacturers refuse them. Occasionally, the crushed silicon cells for metal leaching.
modules are used as a construction material such as a
concrete aggregate. Therefore, many studies explore
physical methods to further process crushed modules Therefore, it can be part of the module disassembly
and separate glass and silicon cells. It is important to stage or the cell recycling stage. For this reason, a
note that none of the physical methods have shown dedicated section to EVA removal is warranted. Tradi-
100% material separation and some of the fluoropolymer tionally there are two approaches to remove EVA from
back sheet remains on silicon cells. silicon modules or cells: pyrolysis or dissolution. It is
reminded that removal of EVA for detachment of silicon
Veolia opened its facility dedicated to silicon module cells from glass does not necessarily lead to 100% ma-
recycling in France in 2018 [16]. After module crush- terial separation as it requires a subsequent step to
ing, they further grind the crushed modules and then separate the detached cells and glass.
use an optical sensor to separate transparent glass and
opaque silicon. However, fine tuning of their optical Pagnanelli et al. [23] used an organic solvent to detach
separation process is needed as their recovered glass silicon cells from glass. Silicon modules were crushed
cullet contains silicon and their recovered silicon cells and sieved. Coarse particles were treated with cyclo-
contain glass. Dias et al. [17] investigated electrostatic hexane at 40e60  C for 1 h. Glass, copper ribbons, sil-
separation of crushed silicon modules, where the icon cells with EVA, and back sheet were largely
separated materials were collected in three pans. They separated. This process has been demonstrated at a pilot
were able to concentrate most glass in one pan, most scale [24]. Azeumo et al. [25] investigated several
polymers in a second pan, and most silicon and some organic solvents (toluene, xylene, 2,4-trimethylpentane,
polymers in a third pan, but not 100%. For example, n-heptane, and N,N-dimethylformamide) for detach-
silver was found in all three pans. Tokoro et al. [18] ment of silicon cells from glass. Silicon modules were
employed selective grinding to separate glass from crushed first. After multiple physical processes such as
polymers. Crushed modules were processed with floatation and magnetic separation, concentrates of
gravity separation to extract silver concentrates, mag- polymers, metals, and glass were obtained. The glass
netic separation for iron concentrates, and sieving to concentrates were treated with organic solvents.
obtain glass concentrates. Glass was liberated from Toluene was found to produce 100% detachment in 1 h
polymers using eccentric grinding. It is noted that at 60  C. To speed up the organic solvent process, Pang
separation of glass from polymers is not necessary if et al. [26] applied microwave and found that the
the glass cullet contains only EVA. When the glass detachment ratio was improved by 10%. They tested
cullet is remelted for new glass products, the EVA several organic solvents (methanol, ethanol, ethylene
vaporizes. glycol, toluene, trichloroethylene, and acetic acid) with
microwave and trichloroethylene was found to be most
Several groups explored high-voltage pulse crushing of effective. Two groups [27,28] applied supercritical CO2
silicon modules for material separation [19e22], which with or without an organic solvent. 100% detachment
is different from the traditional mechanical crushing. A was not achieved so peeling after supercritical CO2
silicon module is placed in water between two elec- treatment was needed.
trodes, and hundreds of high-voltage pulses in excess of
100 kV break the module into particles of different sizes. Fiandra et al. [13] performed pyrolysis to remove EVA
It was found that fewer high-voltage pulses separate the after milling off the fluoropolymer back sheet. The op-
silicon module into glass with silicon cells and back timum conditions for complete pyrolysis of EVA are
sheet, and repeated high-voltage pulses further break 500  C for 1 h. Xu et al. [29] combined organic solvent
the module down to particles [19]. Sieving and some- and pyrolysis to detach largely intact silicon cells from
times floatation separate particles of different sizes and glass. Small silicon modules up to 50  50 mm2 were
densities. Various metals are more concentrated in par- treated with pressurized toluene at 190  C for 2 h. When
ticles of different sizes, but not 100%. It is determined the toluene-treated modules were placed in a furnace at
that high-voltage pulse crushing at 160 kV for 300 pulses 500  C for complete pyrolysis of the EVA and back
results in an energy consumption of 193 J/g [22]. sheet, most silicon cells were extracted intact.

www.sciencedirect.com Current Opinion in Green and Sustainable Chemistry 2023, 44:100863

6 Recycling of solar panels and windmills (2023)

Yan et al. [30] reported potassium hydroxide (KOH) designed for silicon cell recycling, and silicon cells are
plus ethanol to dissolve EVA. This is a new method as just added to the smelter with copper ores. It is
EVA dissolution had been using either an organic solvent reminded that the fluoropolymer back sheet is still
or acid. Small pieces of silicon modules were placed in attached to the silicon cells. Based on conversations with
the basic solution and heated to 200  C for 3 h in an recyclers, the silver recovery rate by smelting is about
autoclave. 100% detachment was observed. This 65%, i.e., about $1.60 worth of silver per module is not
chemistry is more environmentally friendly as it does recovered by smelting.
not emit CO2 and is less hazardous than many organic
solvents. The second approach is chemical processing. It requires
the EVA on the silicon cells to be removed first. Once
An interesting process to separate EVA from silicon cells the cell surface is exposed, metals are leached and
was reported by Li et al. using laser debonding [31]. The recovered first, and then silicon of metallurgical grade or
aluminum frame and junction box were removed first, solar grade is extracted. Nasty chemicals are often
followed by heating the remaining module to peel off required in hydrometallurgy [6e9]:
the back sheet. A laser with a wavelength absorbed by
silicon but not EVA was scanned across the module to  HNO3 is the most common leaching agent for metals.
debond the EVA layer from the silicon cell. The EVA  HF is used to etch off the Si3N4 layer.
layer was then peeled off after a corner of the layer was  Either sodium hydroxide (NaOH) or a mixture of
separated from the silicon cell by scratching it with a HNO3 and HF etches off the emitter and back-
knife. This process should apply to the debonding of the surface field to recover the silicon base as solar
front EVA layer from silicon cells as well since the glass grade silicon (Figure 1(b) and (c)).
is also transparent to the laser wavelength. One question
is about the throughput: how long does it take to process
an entire module? Therefore, mild chemistry and minimization of chemi-
cal waste are two topics of high priority in chemical cell
Table 2 compares the different methods for EVA recycling, along with high recovery rates for metals.
removal on detached silicon cells from glass. Organic Lead must be removed 100% from the chemical waste,
solvents are unlikely to be used in commercial-scale and 100% silver recovery is preferred.
recycling due to their toxicity. Meanwhile, pyrolysis,
base dissolution, and laser debonding are still in the Metal recovery
race. ROSI, a French startup, claims “soft chemistry” to
recover solar-grade silicon and silver from silicon cells
Stage 2: chemical cell recycling [32], although no detail has been provided. Chung et al.
Once silicon cells are detached from glass, cell recycling [33] developed an iodine plus potassium iodide (I2-KI)
can be performed. These cells contain not only the chemistry to leach silver and aluminum from silicon
structures in Figure 1(b) and (c), but also copper rib- cells. Leached aluminum can be precipitated out first
bons with solder. There are two approaches today for cell when the leachate pH is adjusted to 9.6, and silver is
recycling: pyrometallurgical or hydrometallurgical. Sili- recovered with sodium borohydride (NaBH4) as the
con cells can be sent to a copper smelter for the recovery reducing agent. However, the recovered silver contains
of silver and copper, but silicon, lead, and tin are not more impurities than silver: sodium, iodine, and a little
recovered by smelting. The copper smelter is not aluminum. Zante et al. [34] leached silver and
aluminum from silicon cells using redox catalysts.
Aluminum is leached first in an aqueous solution of iron
(III) chloride (FeCl3). Silver is then leached out with
Table 2
copper (II) chloride (CuCl2) in an aqueous solution of
Comparison of methods for EVA removal on detached silicon choline chloride. Under optimized conditions, 95% of
cells from glass. the silver and aluminum is dissolved within 10 min. It is
Method Variations Issues also possible to regenerate the chemicals for reuse in the
leaching process by evaporating the excess water.
Organic solvent  Heating [23,25]  Toxic solvents
 Microwave [26]  Slow process
However, the recovered silver is silver chloride (AgCl),
 Supercritical  Chemical waste which requires further processing to obtain metallic
CO2 [27,28] silver.
Pyrolysis  Pyrolysis only [13]  CO2 emissions
 Organic solvent [29]  High energy input Electrochemistry has also been applied to metal re-
Base dissolution  KOH or NaOH [31]  Slow process
 Chemical waste
covery from silicon cells. Modryzynski et al. [35]
Laser debonding  n/a [31]  Slow process? electrochemically leached silver, copper, and tin from
 Broken modules? silicon cells in sulfuric acid (H2SO4). With a boron-
doped diamond electrode, peroxydisulfate (S2 O28 )
Current Opinion in Green and Sustainable Chemistry 2023, 44:100863 www.sciencedirect.com
 Progress in silicon solar module recycling Tao et al. 7

was produced for metal leaching. Electrowinning metallic lead. Further processing is needed to obtain all
recovered pure metals from the leachate. However, the the lead in the metallic form.
leaching rate is slow and the presence of S2 O28 in-
terferes with silver and tin electrowinning. It is also Silicon recovery
unclear how large the diamond electrode can be for It is uncertain at this time about the quality of silicon
commercial-scale recycling. Deng et al. [36] reported a that recyclers will recover: ferrosilicon, metallurgical-
different electrochemical leaching method, in which a grade silicon, or solar-grade silicon. Recovery of solar-
silicon cell serves as the anode and a silver coil as the grade silicon is preferred as it generates a much higher
cathode. Only the front surface of the silicon cell is in revenue and cuts down the energy consumption signif-
contact with the electrolyte which is silver nitrate icantly as the recovered silicon bypasses the energy-
(AgNO3). Over 95% of the silver is recovered within a intensive Siemens process for silicon purification, but
few minutes. However, this method does not recover it requires a complicated process with a high cost. On
the silver soldering pads on the back of the cell. It may the other hand, there are concerns about the purity of
also be difficult to implement it in commercial-scale the recovered solar-grade silicon not matching virgin
recycling. TG Companies, a startup in Arizona, has silicon, which would degrade the efficiency of the silicon
developed a regenerative chemistry for sliver recovery cell. More importantly, there are several logistic barriers
[37], i.e., the acid for silver leaching is regenerated to the recovery of solar-grade silicon [12].
later in the process during electrowinning and then
reused in silver leaching. While most of the silicon modules on the market employ
p-type silicon wafers, a small portion of them are n-type
Chen et al. [38] reported a leaching, solvent extraction, silicon. This information about module type is often lost
and precipitation process to recover AgCl and hydrox- with end-of-life modules since they were produced 25
ides of lead, copper, and tin (Pb(OH)2, Cu(OH)2, and years ago and the manufacturers are long gone. When p-
Sn(OH)2). They used HCl plus hydrogen peroxide type modules and n-type modules are mixed, the
(H2O2) to selectively leach lead, copper, and tin from recovered silicon contains both p-type and n-type. The
ribbons extracted from silicon modules. To completely mixed p-type and n-type silicon automatically down-
remove the solder and obtain pure copper ribbons, they grades itself to metallurgical-grade silicon [7]. In addi-
attempted two methods: 1) sodium chloride (NaCl) was tion, convincing silicon ingot manufacturers to take in
added to HCl for complete lead leaching and 2) the the recovered solar-grade silicon is a difficult task. They
ribbons were oxidized at 800  C and then polished to want a stable feedstock with consistent quality control,
remove the oxidized solder. The chloride leachate but recyclers have little control over the quantity and
contained three metals, lead, copper, and tin. Tributyl consistency of the incoming end-of-life modules.
phosphate was added first to extract tin, and then
LIX984N was added to extract copper, leaving only lead Nevertheless, the process to recover silicon of different
in the leachate. Ammonium hydroxide (NH4OH) was qualities is summarized here:
added to the three solutions to obtain hydroxides of the
three metals. Silver was recovered through HNO3  The silicon cells detached from glass can be sold
leaching and NaCl precipitation. This process illustrates directly as ferrosilicon without any further processing.
the downside of conventional chemical recycling, i.e., Since the fluoropolymer back sheet is still on the cells,
the large number of chemicals needed and thus large caution must be taken if the cells will go through a
quantities of chemical waste generated. In contrast, high-temperature process.
electrowinning after leaching requires a far fewer  For metallurgical-grade silicon, silicon cells after
number of chemicals. metal recovery are submerged in HF to remove Si3N4
and NaOH to remove aluminum.
There is a non-chemical method to recover silver from  To extract solar-grade silicon, the silicon wafers after
silicon cells [39]. Silicon cells extracted from modules removal of Si3N4 and aluminum are further treated
were placed in water. Electrical contacts were made to with NaOH or a mixture of HF and HNO3 to remove
the copper ribbons and a 15 kV pulse was applied across the emitter and back-surface field.
the copper ribbons causing the silver front electrode to
explode. However, the recovered particles contained
less than 5 wt% of silver. Deng et al. [36] reported a 4N purity for the recovered
silicon. It is possible that the back-surface field was not
Not many studies focused on lead recovery from silicon completely removed, leading to the lower purity. With
cells. Besides HNO3, acetic acid (CH3COOH) plus only KOH, Chen et al. [38] reported a 2N purity. By
H2O2 is an effective and mild leaching agent for lead monitoring the NaOH etching process, TG Companies
[10]. 24-hour electrowinning of the acetic leachate re- was able to recover silicon of a 6N purity [37]. Induc-
sults in 99% lead recovery, but the recovered lead in- tively coupled plasma mass spectrometry and glow
cludes various lead oxides and acetates in addition to discharge mass spectrometry by a third party both
www.sciencedirect.com Current Opinion in Green and Sustainable Chemistry 2023, 44:100863
8 Recycling of solar panels and windmills (2023)

confirmed that all the impurities in recovered silicon, Margolis R, Matsubara K, Metz A, Metzger WK, Morjaria M, Niki S,
Nowak S, Peters IM, Philipps S, Reindl T, Richter A, Rose D,
except the dopant in the base, are below the detection Sakurai K, Schlatmann R, Shikano M, Sinke W, Sinton R,
limit of 0.03 ppm. Stanbery BJ, Topic M, Tumas W, Ueda Y, Van De Lagemaat J,
Verlinden P, Vetter M, Warren E, Werner M, Yamaguchi M,
Bett AW: Terawatt-scale photovoltaics: transform global
Since recovery of solar-grade silicon might be chal- energy. Science 2019, 364:836–838.
lenging and metallurgical-grade silicon generates a low 3. Tao M, Hamada H, Druffel T, Lee J-J, Rajeshwar K: Research
revenue (about 2¢/cell), other more valuable applica- needs for photovoltaics in the 21st century. ECS Journal of
Solid State Science and Technology 2020, 9, 125010.
tions for the recovered silicon are worth pursuing. Zhang
et al. [40] explored a silicon anode for lithium-ion bat- 4. SolarPower Europe: Global market outlook for solar power
2022–2026. https://www.solarpowereurope.org/insights/market-
tery from recovered silicon. After removing all the non- outlooks/global-market-outlook-for-solar-power-2022; 2022.
silicon layers, the silicon wafers were milled and then
5. US Environmental Protection Agency, National Overview: Facts
converted to porous silicon through electrochemical and Figures on Materials, Wastes and Recycling, https://www.
alloying and dealloying with lithium in a molten salt of epa.gov/facts-and-figures-about-materials-waste-and-recycling/
national-overview-facts-and-figures-materials.
lithium chloride and potassium chloride (LiCleKCl).
The porous silicon anode shows good performance with 6. Wang X, Tian X, Chen X, Ren L, Geng C: A review of end-of-life
crystalline silicon solar photovoltaic panel recycling
a capacity of 2427.7 mAh/g at 1 A/g after 200 cycles with technology. Sol Energy Mater Sol Cell 2022, 248, 111976.
a capacity retention rate of 91.5%. 7. Tao M, Fthenakis V, Ebin B, Steenari B-M, Butler E, Sinha P,
Corkish R, Wambach K, Simon ES: Major challenges and op-
portunities in silicon solar module recycling. Prog Photovol-
Summary taics Res Appl 2020, 28:1077–1088.
Recent progress in silicon solar module recycling in both 8. Farrell CC, Osman AI, Doherty R, Saad M, Zhang X, Murphy A,
the research community and private sector is reviewed. Harrison J, Vennard ASM, Kumaravel V, Al-Muhtaseb AH,
Some of the more noticeable progress includes: 1) me- Rooney DW: Technical challenges and opportunities in
realizing a circular economy for waste photovoltaic modules.
chanical milling to remove the fluoropolymer back Renew Sustain Energy Rev 2020, 128, 109911.
sheet; 2) laser debonding of the EVA layer from silicon 9. Deng R, Chang NL, Ouyang Z, Chong CM: A techno-economic
cells; 3) dissolution of EVA with a base; 4) mild chem- review of silicon photovoltaic module recycling. Renew Sus-
istry for silver and lead recovery; and 5) regenerative tain Energy Rev 2019, 109:532–550.
chemistry to minimize chemical waste. These results 10. Click N, Adcock R, Chen T, Tao M: Lead leaching and elec-
* trowinning in acetic acid for solar module recycling. Sol
may lead to practical solutions for several technical Energy Mater Sol Cell 2023, 254, 112260.
barriers in silicon module recycling, facilitating a 90e Lead in silicon solar modules creates an environmental hazard when
95 wt% circularity for silicon modules with an environ- landfilled. Not many papers have investigated lead recovery, and all of
them employ nasty nitric acid. This paper demonstrates a mild chem-
mentally and financially sound recycling process. ical, acetic acid, for lead recovery.
11. US Environmental Protection Agency, End-of-Life Solar Panels:
Declaration of competing interest Regulations and Management, https://www.epa.gov/hw/end-life-
solar-panels-regulations-and-management.
The authors declare the following financial interests/
12. Tao M, Click N, Ricci L: Commentary on technoeconomic
personal relationships which may be considered as po- analysis of high-value, crystalline silicon photovoltaic
tential competing interests: Meng Tao reports a rela- module recycling processes [Solar Energy Materials and
tionship with TG Companies LLC that includes: equity Solar Cells 238 (2022) 111592]. Sol Energy Mater Sol Cell 2022,
239, 111667.
or stocks.
13. Fiandra V, Sannino L, Andreozzi C, Corcelli F, Graditi G: Silicon
* photovoltaic modules at end-of-life: removal of polymeric
Data availability layers and separation of materials. Waste Manag 2019, 87:
97–107.
Data will be made available on request. Previous methods to remove the fluoropolymer back sheet either
require complicated equipment or releases a fluorine exhaust. This
paper demonstrates a simple method for back sheet removal, i.e.,
Acknowledgments mechanical milling.
This work is supported by the US Department of 14. NPC Incorporated, https://www.npcgroup.net/eng/.
Energy through the REMADE Institute, grant no. 21-
15. Ercole P: FRELP 2 project – full recovery end-of-life
01-RR-5014. MT is grateful to Dr. P. Sinha of First Solar photovoltaic,” proceedings of 32nd European photovoltaic
for many insightful discussions. solar energy conference and exhibition. 2016:1775–1783.
Munich.
References 16. Veolia Group: Veolia opens the first European plant entirely
Papers of particular interest, published within the period of review, dedicated to recycling photovoltaic. 2018. https://www.veolia.com/
have been highlighted as: en/newsroom/news/recycling-photovoltaic-panels-circular-
economy-france.
* of special interest
17. Dias P, Schmidt L, Gomes LB, Bettanin A, Veit H, Bernardes AM:
Recycling waste crystalline silicon photovoltaic modules by
1. International Energy Agency: Solar PV. 2022. https://www.iea.org/ electrostatic separation. Journal of Sustainable Metallurgy
reports/solar-pv. 2018, 4:176–186.
2. Haegel NM, Atwater H, Barnes T, Breyer C, Burrell A, Chiang Y- 18. Tokoro C, Nishi M, Tsunazawa Y: Selective grinding of glass to
M, De Wolf S, Dimmler B, Feldman D, Glunz S, Goldschmidt JC, remove resin for silicon-based photovoltaic panel recycling.
Hochschild D, Inzunza R, Kaizuka I, Kroposki B, Kurtz S, Leu S, Adv Powder Technol 2021, 32:841–849.

Current Opinion in Green and Sustainable Chemistry 2023, 44:100863 www.sciencedirect.com

 Progress in silicon solar module recycling Tao et al. 9

19. Akimoto Y, Iizuka A, Shibata E: High-voltage pulse crushing without acid or organic solvent. It employs potassium hydroxide which
and physical separation of polycrystalline silicon photovol- is less hazardous than acid or organic solvent.
taic panels. Miner Eng 2018, 125:1–9.
31. Li X, Liu H, You J, Diao H, Zhao L, Wang W: Back EVA recycling
20. Nevala S-M, Hamuyuni J, Junnila T, Sirviö T, Eisert S, Wilson BP, * from c-Si photovoltaic module without damaging solar cell
Serna-Guerrero R, Lundström M: Electro-hydraulic fragmenta- via laser irradiation followed by mechanical peeling. Waste
tion vs conventional crushing of photovoltaic panels – Manag 2022, 137:312–318.
impact on recycling. Waste Manag 2019, 87:43–50. Previous methods to delaminate the encapsulant are all chemical. This
report demonstrates a physical method to delaminate the encapsulant.
21. Zhao P, Guo J, Yan G, Zhu G, Zhu X, Zhang Z, Zhang B: A novel It cleanly releases the encapsulant layer without any chemical waste,
and efficient method for resources recycling in waste allowing the encapsulant to be recovered.
photovoltaic panels: high voltage pulse crushing. J Clean
Prod 2020, 257, 120442. 32. ROSI Solar, https://www.rosi-solar.com/photovoltaic-modules-
recycling/.
22. Song B-P, Zhang M-Y, Fan Y, Jiang L, Kang J, Gou T-T, Zhang C-
L, Yang N, Zhang G-J, Zhou X: Recycling experimental inves- 33. Chung J, Seo B, Lee J, Kim JY: Comparative analysis of I2-KI
tigation on end of life photovoltaic panels by application of and HNO3 leaching in a life cycle perspective: towards sus-
high voltage fragmentation. Waste Manag 2020, 101:180–187. tainable recycling of end-of-life c-Si PV panel. J Hazard Mater
2021, 404, 123989.
23. Pagnanelli F, Moscardini E, Altimari P, Padoan FCSM, Atia TA,
Beolchini F, Amato A, Toro L: Solvent versus thermal treatment 34. Zante G, Rivera RM, Hartley JM, Abbott AP: Efficient recycling
for glass recovery from end of life photovoltaic panels: * of metals from solar cells using catalytic etchants. J Clean
environmental and economic assessment. J Environ Manag Prod 2022, 370, 133552.
2019, 248, 109313. All previous studies employed nasty chemicals such as nitric acid for
silver recovery from silicon solar cells. This paper demonstrates mild
24. Rubino A, Schiavi PG, Altimari P, Pagnanelli F: Valorization of chemicals to recover silver using of iron chloride, copper chloride, and
polymeric fractions and metals from end of life photovoltaic choline chloride.
panels. Waste Manag 2021, 122:89–99.
35. Modrzynski C, Blaesing L, Hippmann S, Bertau M, Bloh JZ,
25. Azeumo MF, Germana C, Ippolito NM, Franco M, Luigi P, Weidlich C: Electrochemical recycling of photovoltaic mod-
Settimio S: Photovoltaic module recycling, a physical and a ules to recover metals and silicon wafers. Chem Ing Tech
chemical recovery process. Sol Energy Mater Sol Cell 2019, 2021, 93:1851–1858.
193:314–319.
36. Deng R, Dias PR, Lunardi MM, Jia J: A sustainable chemical
26. Pang S, Yan Y, Wang Z, Wang D, Li S, Ma W, Wei K: Enhanced process to recycle end-of-life silicon solar cells. Green Chem
separation of different layers in photovoltaic panel by mi- 2021, 23:10157–10167.
crowave field. Sol Energy Mater Sol Cells 2021, 230, 111213.
37. TG Companies, https://www.tg-companies.com/.
27. Lovato ES, Donato LM, Lopes PP, Tanabe EH, Bertuol DA:
Application of supercritical CO2 for delaminating photovol- 38. Chen W-S, Chen Y-J, Lee C-H, Cheng Y-J, Chen Y-A, Liu F-W,
taic panels to recover valuable materials. J CO2 Util 2021, 46, Wang Y-C, Chueh Y-L: Recovery of valuable materials from
101477. the waste crystalline-silicon photovoltaic cell and ribbon.
Processes 2021, 9:712.
28. Briand A, Leybros A, Doucet O, Vite M, Gasmi A, Ruiz JC,
Lamadie F, Grandjean A: Deformation-induced delamination of 39. Lim S, Imaizumi Y, Mochidzuki K, Koita T, Namihira T, Tokoro C:
photovoltaic modules by foaming ethylene-vinyl acetate with Recovery of silver from waste crystalline silicon photovoltaic
supercritical CO2. J CO2 Util 2022, 59, 101933. cells by wire explosion. IEEE Trans Plasma Sci 2021, 49:
2857–2864.
29. Xu X, Lai D, Wang G, Wang Y: Nondestructive silicon wafer
recovery by a novel method of solvothermal swelling coupled 40. Zhang C, Ma Q, Cai M, Zhao Z, Xie H, Ning Z, Wang D, Yin H:
with thermal decomposition. Chem Eng J 2021, 418, 129457. * Recovery of porous silicon from waste crystalline silicon
solar panels for high-performance lithium-ion battery
30. Yan Y, Wang Z, Wang D, Cao J, Ma W, Wei K, Yun L: Recovery anodes. Waste Manag 2021, 135:182–189.
* of silicon via using KOH-ethanol solution by separating Recovery of solar-grade silicon from silicon solar cells is difficult, and
different layers of end-of-life PV modules. JOM 2020, 72: metallurgical-grade silicon generates little value. This is the first report
2624–2632. on alternative high-value applications for the recovered silicon, as the
Previous methods for encapsulant dissolution require either an acid or anode in lithium ion batteries.
organic solvent. This is the first report on encapsulant dissolution

www.sciencedirect.com Current Opinion in Green and Sustainable Chemistry 2023, 44:100863