Solar Energy Materials & Solar Cells 277 (2024) 113102

Contents lists available at ScienceDirect

Solar Energy Materials and Solar Cells

journal homepage: [Link]/locate/solmat

Crystalline silicon solar cell with an efficiency of 20.05 % remanufactured

using 30 % silicon scraps recycled from a waste photovoltaic module
Jun-Kyu Lee a, 1 , Suk-Whan Ko a , Hye-Mi Hwang a , Woo-Gyun Shin a , Young-Chul Ju a ,
Gi-Hwan Kang b , Hee-Eun Song b , Young-Joo Eo b , Soohyun Bae b , Wolfram Palitzsch c ,
Ingo Röver c , Jin-Seok Lee a, *
a
Renewable Energy System Laboratory, Korea Institute of Energy Research, Daejeon, 34129, Republic of Korea
b
Photovoltaic Laboratory, Korea Institute of Energy Research, Daejeon, 34129, Republic of Korea
c
LuxChemtech GmbH, 09599, Freiberg, Germany

A R T I C L E I N F O A B S T R A C T

Keywords: Several studies have aimed to develop methods for recycling and recovering valuable materials from waste
Photovoltaic modules photovoltaic (PV) modules to keep up with the increasingly stringent waste disposal guidelines. However, silicon
Recycled silicon scraps (Si) recovered from disposed PV modules is rarely used as a raw material in the solar industry. In this study, 30 %
Silicon ingot
Si recovered from waste PV modules was added to a feedstock to grow a 6inch single-crystalline Si ingot using
Solar cell
Efficiency
the Czochralski method. The single-crystalline Si ingot was re-melted and manufactured to produce a higher
quality single-crystalline Si ingot with a minority carrier lifetime of 274–1527 μs and purity of 7N7. Thereafter, a
Si wafer was manufactured via cropping, squaring, and wafering. Finally, the Si wafer was used to fabricate a
solar cell via a conventional process. Solar cell parameters, such as open circuit voltage, short circuit current
density, fill factor, and efficiency, were analyzed using a solar simulator. The manufactured solar cell had an
efficiency of 20.05 %, which is approximately 0.97 % lower than that of commercial wafer-based solar cells.
Moreover, the factors influencing the solar cell efficiency were evaluated.

1. Introduction acetate, and a backsheet. Among these valuable materials, c-Si solar cell
scraps are etched via a chemical treatment process to recover the Si
A report published by the International Energy Agency’s Photovol- scrap. With the advancement of technology, various solar cells such as
taic Power Systems Program (IEA-PVPS) predicts that by 2030, there aluminum-back surface field (Al-BSF), a passivated emitter and rear cell
will be an estimated 1.7–8 million tons of waste photovoltaic (PV) (PERC), tunnel oxide passivated contact (TOPCon), and Heterojunction
modules, which will reach 60–78 million tons by 2050 [1]. If left un- technology (HJT) have been developed. Despite their structural differ-
treated, waste PV modules can harm the environment [2]. Waste PV ences, all these solar cells feature functional layers on the front and rear
modules have already been classified as electrical and electronic surfaces of a Si wafer base. By employing selective etching techniques,
equipment waste (WEEE) by the 2012/19/EU Directive. Moreover, these functional layers can be removed, enabling the recovery of
recycling PV modules has become mandatory in Europe following their high-purity Si [4,5]. According to the International Technology Road-
introduction to individual European countries. According to the revised map for Photovoltaics (ITRPV) report published in 2021, polysilicon and
WEEE guidelines, from 2018, 85 % of waste PV modules in Europe must wafers used to manufacture modules account for 31 % of total produc-
be recovered and 80 % must be recycled [3]. Therefore, numerous tion [6]; therefore, it is crucial to recover high-purity Si and wafers from
technical methods for recycling and recovering valuable materials from waste PV modules. In most studies, PV modules have been separated
PV modules have been reported. A typical c-Si PV module, excluding the using thermal methods, wafers have been etched using chemical pro-
aluminum frame and junction box, comprises a laminated structure cesses, and solar cells have been remanufactured and evaluated [7–13].
composed of glass, a crystalline Si solar cell, a Cu ribbon, ethylene-vinyl However, there are difficulties in handling, including thermal module

* Corresponding author.
E-mail address: jslee@[Link] (J.-S. Lee).
1
Present address: Policy Planning Agency, Chungbuk Technopark, Chengju 28116, Republic of Korea.

[Link]
Received 13 June 2024; Received in revised form 4 August 2024; Accepted 10 August 2024
Available online 16 August 2024
0927-0248/© 2024 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
J.-K. Lee et al. Solar Energy Materials and Solar Cells 277 (2024) 113102

separation or the risk of breakage during chemical solar cell etching recycled Si scraps was 1415 ◦ C, and the growth rate was 0.8 mm/min.
processes, and such difficulties are primarily owing to the thinness of the The rotation velocities of the Si ingot and crucible were 12 and 8 rpm,
solar cells. To be more specific, the thickness of Si used in solar cells respectively. Ar gas (50 lpm) was input during the Si ingot growth
tends to decrease during the chemical etching process, thus increasing process. The first Si ingot cast was used as a raw material for casting the
the risk of breakage during the process. Furthermore, the size of solar second Si ingot to improve the quality of the second ingot; that is, the
cells used in currently recycled modules is typically 4–5 inches [12]; melting and growing processes were performed twice. Once the second
thus, recovered Si wafers cannot be used immediately in commercial Si ingot was grown, 70 % virgin polysilicon (2.1 kg) and 30 % recycled Si
solar cell manufacturing processes. Therefore, it is difficult to directly scraps (0.9 kg) were additionally charged owing to the pot loss during
recover Si wafers from waste modules and reuse them in the solar in- the casting (3 kg) of the first Si ingot. Table 1 lists the proportions of raw
dustry. Accordingly, the most beneficial way to use Si recovered from materials used to manufacture the single-crystalline Si ingots. The total
waste modules is to reuse it as a raw material. In general, solar cells are charge of the first and second ingots was 30 kg. Once the Si ingot was
typically classified into n-type and p-type, with boron (B) or phosphorus grown, wafering was performed using a conventional slurry-based
(P) used as dopants. When Si recovered from solar cells is used as a raw multi-wire sawing process. The electrical characteristics of the ingot
material, it is important to prevent the mixing of n-type and p-type Si to were investigated using a Sinton minority carrier lifetime (MCLT) tester
control dopant concentration like B and P acting as impurities. Con- (BCT-400, Sinton Instruments).
ductivity type (P/N) tester is commonly used to determine the type of Si The concentrations of impurities in CZ single-crystalline silicon, such
in less than 1 s. Therefore, it is essential to implement procedure for as B, P, and heavy metals, were measured using ICP-MS. The concen-
identifying the type of Si with the tester before using Si as a raw material trations of non-metallic impurities in the Si wafer, such as O and C, were
for the manufacturing of solar cells. On the other hand, Si recovered measured using Fourier-transform infrared spectroscopy (FTIR, Nano-
from waste modules is rarely used as a raw material in the solar industry. metrics, QS-1200). A scanning electron microscope (SEM, Hitachi S-
Draoua et al. remanufactured solar cells using Si recovered from waste 4700) and an energy dispersive X-ray spectroscope (EDS, Horiba EMAX
solar cell scrap, and the manufactured cells had the highest efficiency of 7200-H) were employed to observe the microstructure and trace
18.1 % [14]. However, they utilized Si recovered from broken solar cells element of the sliced wafer.
instead of Si recovered from PV modules. By contrast, this study inves-
tigated the manufacturing process and performed a characteristic eval- 2.2. Remanufacturing of PERC solar cells
uation of solar cells made from high-purity single-crystalline Si ingots
using some of the Si scraps recovered from waste. The conventional The solar cell was manufactured on B-doped p-type 156 × 156 mm2
Czochralski (Cz) method was adopted to grow the single-crystalline Si c-Si wafers. The resistivity and thickness of the wafers were 1.6–1.8 Ω
ingot. During the process, Si scraps recovered from PV modules were cm and 280 μm, respectively. The wafers were texturized using a
utilized as a part of the raw material. Subsequently, the electrical mixture of potassium hydroxide (KOH) and isopropyl alcohol solution,
characteristics of single-crystalline Si ingots manufactured by adding 30 etched by approximately 5 μm on each side. Phosphorus oxychloride
% recycled Si are presented. A single-crystalline ingot was processed by (POCl3) was used as a source to form an emitter layer on the wafers.
cropping, squaring, polishing, and wafering. Finally, a solar cell was Following thermal diffusion and phosphocilicate glass removal, the
manufactured via the conventional solar cell manufacturing process, sheet resistance was measured to be 85–90 Ω/sq. Subsequently, the
and its PV performance was evaluated using conventional analysis phosphorus diffusion around the edge of the wafers was removed via
methods. The efficiency of the remanufactured solar cells was compa- edge isolation. An anti-reflectance (AR) coating (SiNx:H) was applied
rable to that of general solar cells, indicating their suitability for appli- using plasma-enhanced chemical vapor deposition. The thickness of the
cation in commercial PV modules. Although recovering high-purity Si AR coating was 800 Å, and it functioned as a surface passivation layer.
from PV modules requires substantial energy, this process consumes The thicknesses of the back surfaces of Al2O3 and SiNx layers were 100
much less energy compared to producing solar cells with raw materials, and 1500 Å, respectively. The screen-printing process was performed to
thereby significantly reducing both energy consumption and carbon form the front and back electrodes. Ohmic contact was formed in an
emissions [15,16]. Numerous studies have already explored the com- infrared lamp-heated belt furnace. Finally, PERC solar cell was manu-
mercial feasibility of recycling PV modules and highlighted the eco- factured, and its electrical characteristics were evaluated. I–V charac-
nomic and environmental benefits of such recycling efforts [17–19]. The teristic curves of the manufactured solar cell were analyzed using a solar
method developed in this study has the large potential to be scaled up for simulator (McScience K210 LAB160). The efficiency of a PERC solar cell
commercial and industrial applications. manufactured using recycled Si wafer (RSW) was evaluated and
compared with that of a solar cell manufactured using virgin polysilicon
2. Experimental procedure wafer (VPW).

Recycled Si scraps were provided by LuxChemtech GmbH, Germany.

Waste modules from the companies Solea and Gamesa Solar were used
as the source for the Si. The chemical etching process for recovering Si
scraps from solar cells comprised only one step, i.e., the simultaneous
Table 1
removal of the front and back electrodes, AR coating, and emitter layer.
Proportion of the raw materials used in single-crystalline Si ingots.
The solar cells were immersed in a solution with a mixture of HNO3 and
HF. They were rinsed with deionized water following the etching pro- Raw material Si ingot Total Content of
charge charge recycled Si scrap
cess. The concentrations of B, P, and heavy metals in the recovered Si of raw material
scraps were measured using inductively coupled plasma mass spec-
Virgin Recycle
trometry (ICP-MS, Thermo Scientific, Element 2).
scrap scrap

1st 21 kg 9 kg 30 kg 30 %
2.1. Casting of single-crystalline Si ingot and wafering –

Si ingot
The single-crystalline Si ingot was grown via the traditional CZ grown
casting method. Accordingly, 70 % virgin polysilicon (21 kg) and 30 % 2nd 2.1 kg 0.9 kg 27 kg 30 kg 30 %
recycled Si scraps (9 kg) were charged in a quartz crucible to grow a Si ingot
grown
single-crystalline Si ingot. The temperature for melting the virgin and

2
J.-K. Lee et al. Solar Energy Materials and Solar Cells 277 (2024) 113102

3. Results and discussion second grown Si ingots, which is shown in Fig. 2. According to the po-
sition, the MCLT of the first grown ingot ranged from approximately
Fig. 1 illustrates the series of processes used to grow the ingots by 150–807 μs; this range is within that of commercially fabricated Si in-
adding recycled Si. To grow single-crystalline Si ingots, the recycled Si gots [23]. It is evident that the MCLT of the top part is higher than that of
scraps were added to virgin polysilicon in a quartz crucible as a part of the bottom part; this is because of the segregation characteristics of
the raw materials. The first ingot grown was then used as the raw ma- metallic impurities in Si. By contrast, the MCLT of the second grown
terial for re-melting and growing the second grown Si ingot. A typical 6″ ingot was confirmed to have improved; it ranged from approximately
Si wafer was manufactured via a conventional wafering process. Finally, 274–1527 μs. In particular, the increase in the MCLT of the bottom part
the conventional solar cell manufacturing process was applied to the was significant, implying that the second grown Si ingot had a higher
wafer of the second grown Si ingot, and a new solar cell was manufac- purity; i.e., it was confirmed that the purity of the Si ingot increased
tured. The PV performance of the 6″ solar cell manufactured using the owing to the re-melting of the Si ingot. During the remelting process, the
second grown Si ingot was evaluated; the solar cell had a high efficiency purity of Si ingot was improved because of segregation and vacuum
of 20.05 %, which was only 0.97 % less than the efficiency of a new refining effect. Various impurities exist in Si, and each of these impu-
wafer-based commercial solar cell (21.02 %). rities exhibits specific characteristics. For example, specific impurities
such as Fe, Ti, Ni, Cr, and Cu were classified as segregation impurities
3.1. Purity of recycled Si scraps owing to their lower segregation coefficient (k). Other impurities such as
Al, Ca, Mg, and Mn have lower vapor pressure than Si, and they were
Table 2 lists the concentrations of major metallic impurities and the classified as volatile impurities [24]. Therefore, the concentration of
calculated purity for Si scraps received from LuxChemtech in Germany. impurities in Si decreased during the remelting process owing to the
The purity of Si scrap was found to be 6N8, which is sufficient to be refining effect, and the MCLT of the second ingot was significantly better
applied to solar-grade Si (SoG Si) feedstock [20]. In particular, Ag, Sn, than that of the first ingot because the purity of Si is a key determinant of
Pb, Cu, and Al contained in the solar module (in the used front and rear the MCLT of Si ingot [25].
electrodes and the Cu ribbon) are rarely detected, as indicated in The concentration of metallic impurities in the bottom part of the
Table 2. This implies that high-purity Si was efficiently recovered from second grown ingot was analyzed, as presented in Table 4. It was
the solar modules. Meanwhile, Fe and Cu levels were found to be slightly confirmed that the concentration of metallic impurities in the Si ingot
higher than their corresponding levels in SoG Si. Coletti [21] reported was lower than that of the recovered Si scraps (Table 2). Most metal
that SoG Si contains Fe and Cu levels of 50 and 10 ppbw, respectively. By impurities were found to be less than 1 ppb; however, the concentrations
contrast, the results of the purity analysis in Table 1 indicate the pres- of Al, Fe, and Ti were found to be 1.02, 2.46, and 7.05 ppb, respectively.
ence of higher Fe and Cu levels. However, Si scrap extracted from a In addition, the concentration of Ti in the bottom part of the Si ingot was
waste PV module can be sufficiently reused in this process because the found to be slightly higher than that of the Si recovered from the PV
recycled Si scraps only accounted for 30 % of the total weight of the module. The Ti concentration of the Si recovered from the PV module
manufactured Si ingot. Meanwhile, the concentrations of P and B in the was 1.67 ppb; however, as indicated in Table 4, the Ti concentration of
recycled Si were 16.2 and 33.1 ppb, respectively, which is satisfactory the bottom part of the Si ingot was 7.05 ppb. As expected, the Si scrap
for SoG Si raw materials [22]. Table 3 lists the specifications of the that had a slightly lower Ti concentration was used in the analysis of the
conventionally manufactured ingot and wafer. The diameter of the recovered Si using ICP-MS. Ti is a slow diffuser at high temperatures and
manufactured ingot was 204 mm, and the length of its body, excluding easily forms oxides; still, it can be incorporated into Si ingots when
the top and tail, was 300 mm. The Si ingot weighed 27 kg, and the pot present in molten Si. Since Ti does not precipitate or getter easily, it can
loss was determined to be 3 kg. The wafer, fabricated via a typical have a detrimental effect on the MCLT [26,27]. The calculated purity of
wafering process, had a length of 156 mm, diagonal of 200 mm, and Si ingot was measured to be approximately 7N8. This value corresponds
thickness of 280 μm. to the bottom part of the Si ingot. Therefore, the overall purity of the
ingot is expected to be higher because the purity of a Cz monocrystalline
ingot increases as the distance decreases from the top. As indicated in
3.2. Characteristics of manufactured ingot and wafer
Table 5, the concentrations of O and C of the Si wafer were measured to
be 9.3 × 1017 and 1.7 × 1016 atoms/cm3, respectively. In general, the
The MCLT was determined based on the positions of the first and

Fig. 1. Overall process of fabricating process of the c-Si solar cell from recycled Si scraps.

3
J.-K. Lee et al. Solar Energy Materials and Solar Cells 277 (2024) 113102

Table 2
Concentration of impurities and calculated purity for recycled Si scraps.
Element (ppb) Purity (%)

Al Ca Mg Mn Fe Ti Ni Cr Cu Ag Sn Pb P B

Recycled Si scrap 7.6 34.9 8.2 0.3 65.4 1.6 1.0 13.5 13.6 1.3 0.5 0.1 16.2 33.1 99.999985 (6N8)

elementary charge. Among these parameters, Jo relies on the doping

Table 3
levels and recombination effect [34]. It implies that the recombination
Specifications of the manufactured ingot and wafer.
of a RSW (Jo = 1.16e-08 A/cm2, n = 1.68) solar cell is larger than that of
Ingot Wafer a VPW (Jo = 3.39e-10 A/cm2, n = 1.36) solar cell because the doping
Diameter Body Weight Length Diagonal Thickness level is almost the same in both solar cells. Meanwhile, the Jsc value of
(mm) length (kg) (mm) (mm) (μm) the RSW solar cell was slightly higher than that of the VPW solar cell.
(mm) More specifically, Jsc depends on the generation and collection of
204 300 27 156 200 280 light-generated carriers in the solar cell; hence, it is influenced by the
optical properties and the collection probability [35]. Additionally, Jsc is
affected by recombination losses because the minority carrier diffusion
length decreases owing to recombination losses. Recombination occurs
at the surface of the bulk of the solar cell. Therefore, the external
quantum efficiency (EQE) was measured to confirm the positions in
which the recombination losses occurred and determine the cause for
the lower Voc and higher Jsc values of the RSW solar cell relative to the
VPW solar cell.

3.4. External quantum efficiency and reflectance of solar cells

Fig. 4 shows the EQEs of the RSW and VPW solar cells. The EQE of

Table 5
Concentration of non-impurities for the second grown Si ingot.
Concentration (atoms/cm3)

Oxygen Carbon

2nd Si ingot grown 9.3 1017 1.7 1016

Fig. 2. Minority carrier lifetimes of the first and second grown Si ingots.

presence of high concentrations of O and C in Si can degrade the effi-

ciency of various defects and solar cells [28–30]. However, the con-
centrations of C and O of the ingot grown in this study are conventional
Cz ingot level [31,32].

3.3. Light I–V curves of solar cells

Fig. 3 illustrates the I–V curves of the RSW and VPW solar cells. The
inset table shown in Fig. 3 indicates the short-circuit current density
(Jsc), open-circuit voltage (Voc), fill factor (FF), and cell efficiency (η). Voc
value of the RSW was lower than those of the VPW solar cell. This
phenomenon is related to the recombination in a solar cell. Generally,
Voc can be expressed using the following equation [33]:
( )
nkT JL
Voc = ln +1 (1)
q Jo

where JL denotes the light-induced current density, Jo indicates the

saturation current density, n is the ideality factor, k symbolizes the
Fig. 3. Light I–V curves of the RSW and VPW c-Si solar cells.
Boltzmann constant, T refers to the temperature, and q indicates the

Table 4
Concentration of impurities and calculated purity for the second grown Si ingot.
Element (ppb) Purity (%)

Al Ca Mg Mn Fe Ti Ni Cr Cu

2nd Si ingot grown 1.02 0.13 0.12 0.03 2.46 7.05 0.06 0.15 0.02 99.999998 (7N8)

4
J.-K. Lee et al. Solar Energy Materials and Solar Cells 277 (2024) 113102

the RSW solar cell was consistently lower than that of the VPW solar cell,
except for the long-wavelength range (1040–1100 nm). Generally, the
EQE depends on surface and bulk recombination, which is affected by
the quality of solar cells, including extrinsic surface defects, quality of
passivation, and metallic impurities. In this case, as both RSW and VPW
solar cells were fabricated via the same conventional process from ingot
to solar cell manufacturing, it is expected that they will have the same
passivation quality and exhibit the same types of surface defects. The
EQE includes the effect of optical losses, such as transmission and
reflection [36]. Hence, the reflectance curves of the RSW and VPW solar
cells are illustrated in Fig. 5. Reflectance measurements confirmed that
the reflectivity of the RSW solar cell was almost the same as that of the
VPS solar cell. On the other hand, a low diffusion length is primarily
affected by the QE at short and intermediate wavelengths, which is
lower for RSW solar cell than that of the VPW solar cell. This is because
of the relatively high metallic impurities detected in the VPW solar cell,
as listed in Table 4. Metallic impurities in a Si wafer decreases the
diffusion length owing to high recombination [37]. In this case, these
results implied the existence of differences in the quality, e.g., differ-
Fig. 5. Reflectance of the RSW and VPW solar cells.
ences in terms of metallic impurities, of the wafer fabricated using
recycled Si and that fabricated using virgin polysilicon. These differ-
ences can be attributed to the presence of Ti in the wafers fabricated by thickness [42]. Consequently, the EQE of a RSW solar cell at long
adding recycled Si. The Ti inserted in Si by diffusion or during Cz ingot wavelengths is larger than that of a VPW solar cell. Additionally, Jsc
growth is electrically active, i.e., the Ti in Si causes a considerable increases with thickness of the wafer because of the longer optical path
reduction in the lifetime [38,39]. length [43]. As mentioned earlier, owing to the differences in the cutting
Consequently, Ti impurities result in a lower diffusion length and process, the thickness of the wafer used to manufacture the RSW solar
higher recombination in the RSW solar cell than those in the VPW solar cell was 280 μm, while the thickness of the wafer used to manufacture
cell. As shown in Fig. 4, the recombination effect caused by metal im- the VPW solar cell was 180 μm; therefore, the Jsc value of the RSW solar
purities present in the solar cell is more pronounced at a short wave- cell was expected to be slightly higher than that of VPW solar cell.
length than at an intermediate wavelength. This is because when light is
incident on the solar cell, most energetic photons are absorbed almost 4. Conclusion
within 1 μm of the wafer. Consequently, most carriers are generated near
the surface [40,41]. In other words, because the proportion of carriers A single-crystalline Si ingot was manufactured using raw materials
formed at the front surface is much higher, the EQE difference between that included Si recovered from a waste PV module. The manufactured
the RSW and VPS solar cells due to recombination is prominent in the ingot was then used to fabricate a solar cell, and its characteristics were
short-wavelength region. evaluated. The purity of Si recovered from the PV module was 6N8,
Meanwhile, it was confirmed that the EQE of the RSW solar cell was indicating that it could be used as a raw material for fabricating Si PV
higher than that of the VPW solar cell at long wavelengths. It can be cells. Moreover, 30 % of recycled Si and 70 % of virgin polysilicon were
attributed to the greater thickness (280 μm) of the wafer used to mixed and melted. Ingots were grown using a commercial Cz process.
manufacture the RSW solar cell, which was approximately 100 μm more The melting and growing processes were repeated to enhance the quality
than the thickness of commercial wafers. The amount of absorbed light of the ingots. The bottom purity of the prepared Si ingot was approxi-
depends on the optical path length, which also depends on the solar cell mately 7N8. Based on the position of the ingot, the MCLT ranged from
274 to 1527 μs. The manufactured ingot was fabricated into a wafer via a
wafering process. The Voc of the RSW solar cell was lower than that of
the VPW solar cell because of the low minority carrier diffusion length
caused by metallic impurities in the wafer. By contrast, the Jsc value was
slightly larger than that of the VPW solar cell owing to the higher
thickness of the wafer. The efficiency of the manufactured wafer-based
solar cell was measured as 20.05 %, which is approximately 0.97 % less
than that of a solar cell manufactured using a commercial wafer.

CRediT authorship contribution statement

Jun-Kyu Lee: Writing – original draft, Formal analysis. Suk-Whan

Ko: Writing – review & editing. Hye-Mi Hwang: Writing – review &
editing. Woo-Gyun Shin: Formal analysis. Young-Chul Ju: Formal
analysis. Gi-Hwan Kang: Writing – review & editing. Hee-Eun Song:
Writing – review & editing. Young-Joo Eo: Writing – review & editing.
Soohyun Bae: Writing – review & editing. Wolfram Palitzsch: Formal
analysis. Ingo Röver: Formal analysis. Jin-Seok Lee: Writing – review
& editing, Supervision, Methodology, Funding acquisition,
Conceptualization.

Declaration of competing interest

Fig. 4. External quantum efficiencies of the RSW and VPW c-Si solar cells. The authors declare that they have no known competing financial

5
J.-K. Lee et al. Solar Energy Materials and Solar Cells 277 (2024) 113102

interests or personal relationships that could have appeared to influence [17] S. Preet, S.T. Smith, A comprehensive review on the recycling technology of silicon
based photovoltaic solar panels: challenges and future outlook, J. Clean. Prod. 448
the work reported in this paper.
(2024) 141661.
[18] M. Tao, V. Fthenakis, B. Ebin, B. Steenari, E. Butler, P. Sinha, R. Corkish,
Data availability K. Wambach, E.S. Simon, Major challenges and opportunities in silicon solar
module recycling, Prog. Photovoltaics Res. Appl. 28 (2020) 1077–1088.
[19] I. D’Adamo, M. Miliacca, P. Rosa, Economic feasibility for recycling of waste
No data was used for the research described in the article. crystalline silicon photovoltaic modules, Int. J. Photoenergy 2017 (2017) 4184676.
[20] J. Safarian, G. Tranell, M. Tangstad, Processes for upgrading metallurgical grade
Acknowledgments silicon to solar grade silicon, Energy Proc. 20 (2012) 88–97.
[21] G. Coletti, Impurities in Silicon and Their Impact on Solar Cell Performance,
Utrecht University, Utrech, 2011.
This research was financially supported by the Ministry of Trade, [22] S. Pizzini, Towards solar grade silicon: challenges and benefits for low cost
Industry, and Energy (MOTIE) and Korea Institute for Advancement of photovoltaics, Sol. Energy Mater. Sol. Cells 94 (2010) 1528–1533.
[23] J.E.A. Maurits, H.J. Dawson, C.H. Weaver, The effect of polysilicon impurities on
Technology (KIAT) through the International Cooperative R&D program minority carrier lifetime in Cz silicon crystals. The Conference Record of the
(P0005441). Further support came from the Korea Institute of Energy Twenty-Second IEEE Photovoltaic Specialists Conference-1991, IEEE, 1991.
Technology Evaluation and Planning (KETEP) and the Ministry of Trade, [24] J.K. Lee, J.S. Lee, B.Y. Jang, J.S. Kim, Y.S. Ahn, G. H Kang, H.E. Song, M.G. Kang,
C.H. Cho, 6″ crystalline silicon solar cell with electron-beam melting-based
Industry & Energy (MOTIE) of the Republic of Korea (RS-2023- metallurgical route, Sol. Energy 115 (2015) 322–328.
00303745, 20213030010390). [25] A. Cuevas, M. Stocks, D. McDonald, M. Kerr, C. Samundsett, Recombination and
trapping in multicrystalline silicon, IEEE Trans. Electron. Dev. 46 (10) (1999)
2026–2034.
References
[26] J. Hofstetter, J.F. Lelievre, C. del Canizo, A. Luque, Acceptable contamination
levels in solar grade silicon: from feedstock to solar cell, Mater. Sci. Eng. B 159
[1] S. Weckend, A. Wad, G.A. Heath, End-of-life Management: Solar Photovoltaic (2009) 299–304.
Panels, National Renewable Energy Laboratory (NREL), Golden, CO (United [27] G. Coletti, P.C. Bronsveld, G. Hahn, W. Wart, D. Macdonald, B. Ceccaroli,
States), 2016. No. NREL/BK-6A20-66178. K. Wambach, N.L. Quang, J.M. Fernandez, Impact of Metal contamination in
[2] J. Tao, S. Yu, Review on feasible recycling pathways and technologies of solar silicon solar cells, Adv. Funct. Mater. 21 (5) (2011) 879–890.
photovoltaic modules, Sol. Energy Mater. Sol. Cells 141 (2015) 108–124. [28] S.K. Pang, A. Rohatgi, Effect of oxygen concentration on lifetime in magnetic
[3] E.C. Directive, Directive 2012/19/EU of the European Parliament and of the Czochralski silicon, J. Electrochem. Soc. 138 (2) (1991) 523–527.
Council of 4 July 2012 on waste electrical and electronic equipment (WEEE), Off. [29] P.C. Bronsveld, P. Manshanden, F.O. Lenzmann, Ø. Gjerstad, E.J. Øvrelid,
J. Eur. Union L197 (2012) 38–71. Recycling of p-type mc-Si top cuts into p-type mono c-Si solar cells, Energy Proc. 38
[4] S. Gao, X. Chen, J. Qu, T. Gau, H. Shi, F. Pang, L. Guo, X. Qu, D. Wang, H. Yin, (2013) 536–541.
Recycling of silicon solar panels through a salt-etching approach, Nat. Sustain. 7 [30] H.J. Möller, L. Long, M. Werner, D. Yang, Oxygen and carbon precipitation in
(2024) 920–930. multicrystalline solar silicon, Phys. Status Solidi A 171 (1) (1999) 175–189.
[5] Z. Zhang, M. Liu, L. Wang, T. Chen, L. Zhao, Y. Hu, C. Xu, Optimization of indium [31] S. Richter, T. Kayser, R. Kunert, S. Wahl, P. Dold, C. Hagendorf, Development and
recovery from waste crystalline silicon heterojunction solar cells by acid leaching, Investigation of Long-Term Stable Quartz Glass Crucible for the Czochralski
Sol. Energy Mater. Sol. Cells 230 (2021) 111218. Crystallization of Silicon, Fraunhofer CSP, 2015, pp. 1–9. Research Report.
[6] V.P. Equipment, International Technology Roadmap for Photovoltaic (ITRPV), [32] G. Kissinger, Oxygen impurity in crystalline silicon. Handbook of Photovoltaic
2021. Silicon, Springer, Berlin, Heidelberg, 2019, pp. 399–436.
[7] J. Shin, J. Park, N. Park, A method to recycle silicon wafer from end-of-life [33] O. Von Roos, P.T. Landsberg, Effect of recombination on the opencircuit voltage of
photovoltaic module and solar panels by using recycled silicon wafers, Sol. Energy a silicon solar cell, J. Appl. Phys. 57 (1985) 4746–4751.
Mater. Sol. Cells 162 (2017) 1–6. [34] P. Procel, V. Maccaronio, F. Crupi, G. Cocorullo, M. Zanuccoli, P. Magnone,
[8] E. Klugmann-Radziemska, P. Ostrowski, Chemical treatment of crystalline silicon C. Fiegna, Analysis of the impact of doping levels on performance of back contact –
solar cells as a method of recovering pure silicon from photovoltaic modules, back junction solar cells, Energy Proc. 55 (2014) 128–132.
Renew. Energy 35 (2010) 1751–1759. [35] A.A.E. Ibrahim, M.R.I. Ramadan, S. Aboul-Enein, A.A.A. ElSebaii, S.M. El-
[9] E. Klugmann-Radziemska, P. Ostrowski, K. Drabczyk, P. Panek, M. Szkodo, Broullesy, Short circuit current Isc as a real non-destructive diagnostic tool of a
Experimental validation of crystalline silicon solar cells recycling by thermal and photovoltaic modules performance, Int. J. Renew. Energy Resour. 1 (3) (2011)
chemical methods, Sol. Energy Mater. Sol. Cells 94 (2010) 2275–2282. 162–168.
[10] L. Frisson, K. Lieten, T. Bruton, K. Declercq, J. Szlufcik, H. De Moor, M. Goris, [36] A. Nakane, H. Tampo, M. Tamakoshi, S. Fujimoto, K.M. Kim, S. Kim, H. Shibata,
A. Benali, O. Aceves, Recent improvements in industrial PV module recycling, in: S. Niki, H. Fujiwara, Quantitative determination of optical and recombination
Proceedings of the Sixteenth European Photovoltaic Solar Energy Conference, losses in thin-film photovoltaic devices based on external quantum efficiency
Routledge, Glasgow, UK, 2000, pp. 2160–2163. analysis, J. Appl. Phys. 120 (2016) 064505.
[11] J. Park, W. Kim, N. Cho, H. Lee, N. Park, An eco-friendly method for reclaimed [37] K. Bothe, R. Sinton, J. Schmidt, Fundamental boron-oxygen-related carrier lifetime
silicon wafers from a photovoltaic module: from separation to cell fabrication, limit in mono- and multicrystalline silicon, Prog. Photovoltaics Res. Appl. 13
Green Chem. 18 (6) (2016) 1706–1714. (2005) 287–296.
[12] J.K. Lee, J.S. Lee, Y.S. Ahn, G.H. Kang, H.E. Song, J.I. Lee, M.G. Kang, C.H. Cho, [38] J.W. Chen, A.G. Milnes, A. Rohatgi, Titanium in silicon as a deep level impurity,
Photovoltaic performance of c-Si wafer reclaimed from end-of-life solar cell using Solid State Electron. 22 (1979) 801–808.
various mixing ratios of HF and HNO3, Sol. Energy Mater. Sol. Cells 160 (2017) [39] G. Coletti, Sensitivity of state-of-the-art and high efficiency crystalline silicon solar
301–306. cells to metal impurities, Prog. Photovoltaics Res. Appl. 21 (5) (2013) 1163–1170.
[13] J.K. Lee, J.S. Lee, Y.S. Ahn, G.H. Kang, H.E. Song, M.G. Kang, C.H. Kim, C.H. Cho, [40] N. Guerra, M. Guevara, C. Palacios, F. Crupi, Operation and physics of photovoltaic
Simple pretreatment processes for successful reclamation and remanufacturing of solar cells: an overview, I+ D Tecnológico 14 (2) (2018) 84–95.
crystalline silicon solar cells, Prog. Photovoltaics Res. Appl. 26 (2018) 179–187. [41] S.R. Wenham, M.A. Green, Silicon solar cells, Prog. Photovoltaics Res. Appl. 4
[14] A.D. Draoua, B. Martel, S. Dubois et al., On the fabrication of solar cells based on (1996) 3–33.
newly produced recycled silicon. In Proceedings of the 33rdEur. Photovolt. Sol. [42] K. Amri, R. Belghouthi, M. Aillerie, R. Gharbi, Device optimization of a lead-free
Energy Conf. Exhib., n.d., p. 483–487. perovskite/silicon tandem solar cell with 24.4% power conversion efficiency,
[15] R. Deng, N.I. Chang, Z. Ouyang, C.M. Chong, A techno-economic review of silicon Energies 14 (12) (2021) 3383.
photovoltaic module recycling, Renewable Sustainable Energy Rev. 109 (2019) [43] C.J.J. Tool, A.R. Burgers, P. Manshanden, A.W. Weeber, B.H.W. van Straaten,
532–550. Influence of wafer thickness on the performance of multicrystalline Si solar cells:
[16] A. Mulazzani, P. Eleftheriadis, S. Leva, Recycling c-Si PV modules: a review, a an experimental study, Prog. Photovoltaics Res. Appl. 10 (4) (2002) 279–291.
proposed energy model and a manufacturing comparison, Energies 15 (2022)
8419.