High Recovery of Selenium from Kesterite-Based Photovoltaic Cells

Article abstract of DOI:10.1002/ejic.202000261

The use of photovoltaic cells is constantly increasing and, in particular, a new generation of thin-film photovoltaic (PV) cells is under development. The absorber of these new cells, kesterite (CZT(S)Se), is composed of abundant chemical elements. Nonetheless, the development of the recycling process for these elements is indispensable for circular economy. This research is focused on the recovery of selenium by thermal oxidation and subsequent reduction. Thus, recycling of selenium has been firstly studied on synthetic kesterite and then validated in a real sample of kesterite extracted from glass-based PV cells. The best results were obtained in a vertical tubular furnace at 750 °C with an input of 20 mL/min of air. The posterior reduction process of selenium oxide was achieved by ascorbic acid, a common and economic reagent. Real kesterite was extracted from PV cells by thermal treatment at 90 °C for 1 hour to remove the encapsulant and ulterior treatment with HCl for the release of kesterite absorber. Optimal conditions from synthetic kesterite were applied to a real sample, recovering more than 90 % of selenium with a purity of 99.4 %.

Full text of DOI:10.1002/ejic.202000261

European Journal of Inorganic Chemistry
Accepted Article
Title: High Recovery of Selenium from Kesterite-based Photovoltaic
Cells
Authors: Maria Pilar Asensio, Elisa Abas, Jose Luis Pinilla, and
Mariano Laguna
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Inorg. Chem. 10.1002/ejic.202000261
Link to VoR: https://doi.org/10.1002/ejic.202000261
0
1/2020

European Journal of Inorganic Chemistry
10.1002/ejic.202000261
FULL PAPER
High Recovery of Selenium from Kesterite-based Photovoltaic Cells
Maria Pilar Asensio,[a],[b] Elisa Abas, Jose Luis Pinilla, Mariano Laguna,*, [a]
[a]
[c]
.
[
a]
Dr. M.P. Asensio, Dr. E. Abás, Dr. M. Laguna*
Departamento de Química Inorgánica
Instituto de Síntesis Química y Catálisis Homogénea, Universidad de Zaragoza-CSIC
Plaza S. Francisco s/n, 50009, Zaragoza, Spain
E-mail: m.laguna@unizar.es, url http://www.isqch.unizar-csic.es/ISQCHportal/grupos.do?id=42
Dr. M.P. Asensio
[
b]
Weee Internacional Recycling S.L., P. I. San Miguel, Isaac Newton, 4, 50830, Villanueva de Gállego, Spain
[
c]
Dr. J.L. Pinilla
Instituto de de Carboquímica, CSIC
C/ Miguel Luesma Castán, 4, 50018, Zaragoza, Spain
Abstract: The use of photovoltaic cells is constantly increasing and
earth-abundant and non-toxic elements.[5] Selenium in traces is
essential to human health; however, in a higher concentration of
400 µg/day is harmful.[ The most important system that selenium
affects is terrestrial, since it is related to agricultural activities and
the food chain.[7] Therefore, it is imperative to develop a recycling
process for this chemical element.
in particular, a new generation of thin-film photovoltaic (PV) cells is
under development. The absorber of these new cells, kesterite
6]
(CZT(S)Se), is composed of abundant chemical elements.
Nonetheless, the development of the recycling process for these
elements is indispensable for circular economy. This research is
focused on the recovery of selenium by thermal oxidation and
subsequent reduction. Thus, recycling of selenium has been firstly
studied on synthetic kesterite and then validated in a real sample of
kesterite extracted from glass-based PV cells. The best results were
obtained in a vertical tubular furnace at 750 ºC with an input of 20
mL/min of air. The posterior reduction process of selenium oxide was
achieved by ascorbic acid, a common and economic reagent. Real
kesterite was extracted from PV cells by thermal treatment at 90 ºC
during 1 hour to remove the encapsulant and ulterior treatment with
HCl for the release of kesterite absorber. Optimal conditions from
synthetic kesterite were applied to a real sample, recovering more
than 90% of selenium with a purity of 99.4%.
Critical Raw Materials (CRM) are materials economically and
strategically important for the European economy but there is a
risk in their supply, like indium and gallium.[8] Development of
CRM – free technology ensures a sustainable and circular
economy, and that is why CZT(S)Se thin-film PV devices are
great candidates as an alternative to actual solar panels.[
Besides, CZT(S)Se thin-film photovoltaic devices produce less
energy consumption and low production costs than the
conventional PV devices.[
9]
10]
The Circular Economy represents a step ahead of wastes
recovery. It requires one Inverse Synthesis of the materials to
obtain the elements/compounds that were used in their synthesis,
ready to be reused for the same function.
Process for the obtaining and purification of selenium was already
patent in 1946 by Clack and Elkin, in which selenium was burnt in
a stream of pure oxygen to obtain selenium dioxide.[11] This
Introduction
oxidation reaction was carried out in gas phase of selenium, with
the presence of a catalyst and selenium dioxide was recovered
by condensation. However, a re-sublimation step was required to
remove impurities from the catalyst. This process was also
applicable to recover selenium from anode slimes and other
selenium-bearing materials as mud, sludge, dust, etc.[12] Then,
selenium dioxide could be reduced to elemental selenium by
several reducing agents such as aldehydes and ketones, sulphur
dioxide gas, or inorganic reducing agents like iodide, stannous
Global energy consumption is continuously rising, and therefore
renewable energy resources have become important. The Sun is
a source of sustainable energy that produces sunlight, which can
be converted directly to electricity by photovoltaic (PV) devices.[1]
It is expected that the increase of photovoltaic energy in the global
market would result in a growth of waste panels reaching up to
2]
5
.5-6 million tons for 2050.[ Therefore, there is an imperious
need to develop a recycling process for PV panels.
Recycling of solar panels has been studied by different methods
such as dissolution, ultrasonic irradiation, electro-thermal heating,
pyrolysis, mechanical separation and physical disintegration,
among others. However, most of them are still on research, and
the commercialized processes recover only a small portion of the
materials. This issue is mainly due to profitability, since it is more
economical to dispose of solar panels in landfill than recycle
them.[3]
compounds, ferrous salts, hydrazine, hydroxylamine among
[13] [14]
others.
-
Currently, there is still no recycling process developed for
CZT(S)Se PV cells, while there is one for recycling CIGS PV cells.
In this absorber, zinc and tin have been substituted for indium and
gallium, and selenium has been introduced. In this CIGS PV,
metals were recovered by thermal oxidation and solvent
[15]
extraction or electrodeposition.
Selenium could be recovered
There is a large variety of solar panels in terms of structure
by several processes such as sulphating roast, soda roast,
oxidizing roast or chlorination processes.[16] In the case of CIGS
material, the absorber was submitted to thermal oxidation for
separation of selenium dioxide at 800 ºC during 1 hour in an
oxygen-containing atmosphere.[17]
(
mono-crystalline silicon or amorphous silicon); and also in terms
of composition (Cu(InGa)Se (CIGS), cadmium telluride (CdTe),
dye-sensitized, organic and multi-junction).[4] Currently, kesterite
absorbers, Cu ZnSnS (CZTS) and Cu ZnSn(S,Se) (CZTSSe),
have received interest, especially since they are composed of
2
2
4
2
4
1
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.202000261
FULL PAPER
The starting point of recycling CZT(S)Se is the recovery of
selenium for its unlike chemical properties regarding the other
metals.[18] In this work, a process has been developed for the
recovery of selenium by two main steps: thermal oxidation and
reduction. Thermal oxidation of kesterite separated selenium from
copper, zinc and tin; and reduction provided elemental selenium
from its oxidized form. In the first place, kesterite was synthesized
for the development and optimization of the recycling process and
then, the process was applied to real kesterite extracted from
glass-based PV cells for validation of the process.
.
Results and Discussion
Development and optimisation of the process
Kesterite was synthesized for the development and optimization
of the recycling process of selenium from kesterite PV cells.
Figure 1. Raman spectroscopy of synthetic CZTSSe.
Deeper structural studies based on X-Ray Diffraction allowed to
identify kesterite, CZTSSe, as the main phase present,[
19b]
and as
Synthesis and characterisation of kesterite
secondary phases Zn(S,Se) and Sn(S,Se) were recognised,
Kesterite powder, Cu
2 4
ZnSn(S,Se) , was synthesized in ampoules
although only in a few percentage (Figure 2).
by mixing its constituents according to a previously described
[
10]
method by Shah and coworkers,
although other synthetic
routes have been reported.[19] Kesterite was obtained in form of a
black mineral, easily pulverized manually. Composition and
proportion of the chemical elements were confirmed by XRF
analysis, and molar ratio was calculated in relation to 1 mol of zinc
(
Table 1, Figures S1 and S2).
Table 1. AnalyS1sis of XRF of synthetic kesterite.
Result
Cu
25.7
1.9
Zn
Sn
Se
Composition (%)
Ratio
13.8
1.0
1.0
25.9
1.0
1.0
34.6
2.1
2.0
Figure 2. X-Ray Diffraction of synthetic CZTSSe.
Theoretical ratio
2.0
[
21]
Lattice parameters were determined by Le Bail refinement,
and
S/(S+Se) ratio from Vegards law.[ CZTSSe: a=5.564 (1) Å,
c=11.089 (2) Å, S/(S+Se)≈0.5; and for secondary phases,
Zn(S,Se): a=5.465(1)Å, S/(S+Se)≈0.77, and Sn(S,Se):
S/(S+Se)≈0.1.
22]
Due to the fact that sulphur cannot be analysed by this technique,
the presence of sulphur in kesterite structure was confirmed by
Raman spectroscopy. As it can be seen in Figure 1, Raman
spectrum is modified when sulphur is incorporated in the mixture,
and additional peaks arise as a consequence of the S-Se vibration
Thermal oxidation of kesterite
-1
(
peaks at 190, 219 and 329 cm ). At the same time, the vibrations
Once kesterite was synthesized and characterized, it was
submitted to thermal treatment in different piece of equipment for
oxidation and extraction of selenium.
of Se-Se from CZTSe disappeared. Moreover, an assessment of
sulphur content could be performed by using a calibration
method,[ and for this sample was estimated to be [S]/([S] + [Se])
20]
=
0.49.
Muffle
Synthetic kesterite was submitted to thermal treatment in a muffle
and white crystals of selenium dioxide were condensed at the exit
of this piece of equipment. Results of the residual sample of
kesterite thermally oxidized in the muffle are presented in Table
2, and it was concluded that selenium extraction was achieved at
temperatures above 600 ºC.
2
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.202000261
FULL PAPER
Table 2. Extraction of selenium by thermal oxidation of synthetic
copper, zinc and tin remained in the residual sample oxidized
kesterite in a muffle.
(Figure 3).
Temperature (ºC)
400
900
500
900
600
240
700
30
800
20
Time of reaction (minutes)
Extraction of Selenium (%) 61.5 88.2 98.4 99.4 99.2
It was observed that selenium separation started at temperatures
above 400 ºC and an increase of temperature resulted in a
decrease of the reaction time required and moreover, an
improvement of selenium extraction from kesterite. Taking into
account time and temperature parameters, optimal reaction
conditions were established at 700-800 ºC and 20-30 minutes for
thermal treatment of 1.0 g of synthetic kesterite.
Figure 3. Extraction of selenium from kesterite by thermal
oxidation with an input of air.
Apparently, extraction of selenium was improved at higher air
flows (60 vs 20 mL/min), even though the recovery percentage
decreased (Table 4). The glass wool at the exit of the quartz
reactor appeared with coloured powder, showing that high air
flows produced a loss of selenium through the system. Therefore,
optimal reaction parameters were set to 20 mL/min of airflow and
a temperature of 750 ºC to maximized selenium extraction and
recovery (Figure S3). Differences between extraction and
recovery results were attributed to the experimental design for the
recovery of the condensed selenium.
Once selenium was extracted, the residual sample staid for three
more hours in the muffle, and the sample was recovered with the
following proportion Cu:Zn:Sn 2.0:0.9:1.1. This reveals that
reaction conditions were appropriate to extract solely selenium,
since more time of reaction did not produce extraction of the other
metals present. However, we were not able to collect efficiently
the extracted selenium in this piece of equipment.
Small capacity furnace
Because of the problems observed during the muffle method for
the total recovery of selenium, an induction furnace was
considered as an alternative to carrying out the thermal oxidation.
In this case, longer reaction time was required and selenium was
not completely extracted from the residual sample. Based on
Table 3, an increase of temperature not only did not imply a better
extraction of selenium, but also resulted in tin extraction.
Furthermore, the residual sample was recovered as a metal alloy
instead of the powder recovered from the muffle, preventing its
facile manipulation.
The condensed powder had a reddish coloration that was
attributed to an allotrope of selenium, [26] suggesting that
selenium was recovered in elemental state as well as oxidized.
This was confirmed by ICP-OES analysis, which revealed that
composition of condensed selenium was of approximately 49%
2
SeO and 51% Se. Initially, the presence of elemental Se was
attributed to the presence of sulphur which was, as well as the
rest of element present, oxidized to sulphur dioxide and, in this
form could reduce selenium dioxide to selenium.[
12]
This
hypothesis was checked by thermal oxidation of kesterite without
sulphur, Cu ZnSnSe ; in this case, selenium condensed had the
same reddish appearance and it was composed again for a
mixture of selenium and selenium dioxide, according to ICP-OES
analysis. It was concluded that the mechanism of selenium
extraction was not only thermal oxidation, but also evaporation.
Table 3. Extraction of Se and Sn during thermal treatment of
kesterite in the small capacity furnace.
2
4
Temperature (ºC) 700
750
17.1 46.5 33.4 14.2 11.6
61.7 100
800
900
1000
%
%
Se extraction
Sn Extraction
0
0
0
Table 4. Extraction of selenium by thermal oxidation of synthetic
kesterite in a tubular furnace.
Loss of tin was attributed to formation of tin sulphide, which was
decomposed and evaporated.[23] The different results obtained
were attributed to the change in the air capacity in the used pieces
of equipment, muffle and furnace. Muffle had 10 times more
volume than the crucible inside the induction furnace, having
more air capacity and therefore, more oxygen available for
oxidation. These results suggested that it is needed a temperature
range of 700-750 ºC and oxygen for oxidation and extraction of
selenium.
T
(ºC)
Air Flow
Final composition of kesterite %Se
sample
Se (%)
(
mL/min)
Extraction Recovery
%
Cu %Zn %Sn %Se
7
7
7
7
7
00 40
00 60
30 40
30 60
50 20
37.5
37.8
38.1
37.6
38.3
21.1
21.5
23.6
23.8
23.7
33.2
35.8
37.8
38.2
37.7
8.2
4.9
0.5
0.4
0.3
77.3
86.7
98.7
99.0
99.3
72.4
69.6
80.0
72.1
94.4
Tubular furnace
The importance of an optimal oxygen concentration was taking
into account, and therefore a tubular furnace was considered. The
great advantage of this furnace is the possibility of an airflow input,
and consequently the presence of oxygen. In this experiment,
thermal oxidation of kesterite produced selenium condensed in
the upper zone of the quartz reactor for its recovery, whereas
3
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.202000261
FULL PAPER
Reduction of selenium dioxide
from the real kesterite sample. Finally, reduction with ascorbic
acid provided elemental selenium of 99.4% purity, according to
XRF and ICP-OES analysis.
The resulting selenium dioxide must be transformed into metallic
selenium in order to be re-introduced in PV cells or other
applications. Selenium dioxide was recovered from the tubular
furnace by dissolution, washing the quartz reactor with distilled
water and then, reduced for obtaining elemental selenium. As it
Further works are in progress in our lab and in Weee
Internatinonal Recycling S.L. to scale up the process, quantifying
well these common reagents used and evaluating the
environmental pollution assessment. The recovery of selenium,
despite not been a critical raw material, it is compulsory in terms
of environmental pollution, because Se is harmful in big
concentration of 400 µg/day.
was mentioned above, Riley reduction method was performed
with 2-phenylacetophenone and 4-acetylbiphenyl,[17] but the
obtained results were unsuccessful. It was found that 2-
phenylacetophenone provided low yield and purity of selenium
reduced by 4-acetylbiphenyl was relatively poor because of
precipitation of the reducing agent oxidized. Economic aspects
must be considered, and both reducing agents were not
economically viable comparing to alternative reducing agents
studied (Table 5).
Conclusion
Table 5. Results of selenium dioxide reduction by several
reducing agents.
A process has been developed for the recovery of selenium from
kesterite by thermal oxidation and reduction. This process offers
advantages in terms of costs during thermal oxidation with air
instead of oxygen, and by using economic reducing agents during
reduction, using ascorbic acid instead of Riley reagents.
Furthermore, this process has been validated in a real sample of
kesterite recovered from glass-based PV cells. Kesterite absorber
was recovered after delaminating the encapsulant, by thermal
treatment and chemical treatment with hydrochloric acid. This
process was performed with common chemical reagents
providing the recovery of more than 90% of selenium from
kesterite with a purity of 99.4%. Furthermore, the recycling
process developed is viable for the scale-up to an industrial
process.
Purity of
Yield of
Reducing agent
98% purity)
Price*
Price*
selenium
reduction
%)
(
(€/100 g)
(€/1 mol Se)
(
%)
(
2
-
132
259
68.9
99.2
Phenylacetophenone
4
-Acetylbiphenyl
64
36
33
21
125
47
95.1
17.0
51.2
95.6
50.5
99.8
32.8
99.4
Zinc
Tin chloride
Ascorbic acid
125
74
a
https://cymitquimica.com/
Based on the results of Table 5, zinc powder was the most
economic reducing agent and gives the best purity (Figure S4). ,
although the yield of selenium reduction was really poor. Tin
chloride offered better results in terms of yield but lowers,
regarding purity. Ascorbic acid resulted as the optimal reducing
agent not only for economic aspects but also for results of purity
and yield (Figure S5).
Experimental Section
Chemical reagents used for the recovery of selenium were purchased from
Sigma-Aldrich (Madrid, Spain) and used as received. Kesterite glass-
based PV cells were provided by IREC (Institut de Recerca en Energia de
Catalunya) and IMRA Europe.
Validation of the process on kesterite from PV cells
Experimental measurements and equipment
All these experiments allowed us to optimize the conditions of the
selenium recovery process, but they must be applicable to a real
recycling process of glass-based PV cells.
X-Ray Fluorescence analysis were carried out in a Fischerscope XRAY
XAN, with WinFTM software using measured conditions of HV:50kV,
PrimFe, Al 1000; and also in a Thermo Electron, ARL ADVANT´XP using
UNIQUANT analytic software.
Kesterite absorber was recovered from glass-based PV by
thermal treatment (to disassemble from the plastic protective
encapsulation), followed by a chemical treatment with HCl 37%.
Thermal treatment of PV cells allowed manual separation of the
cover plastic with the filler, leaving the PV cell delaminated. HCl
treatment of the resulting cell, dissolved the front layers of CdS,
ZnO, ITO, and In; molybdenum stayed stuck to glass substrate
and kesterite was released. Kesterite remained insoluble
suspended in solution and it was recovered by filtration, dried, and
analysed by XRF. A 92.5% purity was obtained, being the
impurities 6.3% Mo and 1.2% In.
Inductively Coupled Plasma Optical Emission Spectroscopy was done in
a Thermo Elemental IRIS intrepid with an automatic injector and 700 K of
particle temperature.
Raman Spectroscopy analysis was carried out in a LabRam HR800-UV
and T64000 Horiba-Jobin Yvon spectrometers.
X-Ray Diffraction measurement was performed in
a
RIGAKU
Kesterite from PV cells was submitted to thermal oxidation
according to the optimal conditions in a vertical tubular furnace.
The impurities of Mo and In remained in the residual sample,
without affecting recovery of selenium. Analysis of the residual
sample showed that more than 99% of selenium was extracted
diffractometer “D-Max/2500” with rotator anode. Measurements
a
conditions were 40 kV and 80 mA using a copper anode and a graphite
monocromator.
4
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.202000261
FULL PAPER
Muffle thermal oxidation of kesterite was done in an Ashes Furnace FFC /
A Italimpianti Orafi.
Figure 4. Experimental set up for thermal oxidation of kesterite in an
induction furnace.
Small capacity furnace thermal treatment of kesterite was performed in a
Schmelzofen Goldbrunn Therm smelting furnace model 3000.
Tubular furnace
Thermal oxidation in tubular furnace was performed in vertical
configuration.[24] The quartz reactor inside the furnace has 759 mm length,
15 mm inner diameter and 18 mm external diameter. Tubular furnace has
half length of the quartz reactor and therefore, part of the reactor is at room
temperature for condensation and therefore, recovery of selenium.
Kesterite sample (2.0 g) was placed in the middle of the reactor inside the
furnace and an air input from the bottom. Vertical configuration favoured
contact between sample and oxygen from the air. Inside quartz reactor,
glass wool was placed at three different heights (Figure 5):
Tubular furnace kesterite was thermally oxidized in a Watlow ceramic
furnace heated with an electric furnace and with an Iberfluid mass flow
controller.
Development and optimisation of the process
Synthesis and characterisation of kesterite
Kesterite powder was synthesized in quartz ampoules of 20 cm length and
1
.5 cm diameter. In each ampoule, 10.0 g of kesterite were prepared by
1.
2.
Centre of the furnace, where temperature is maximum,
to hold the sample.
mixing its constituent metals in the following ratio Cu:Zn:Sn:S:Se 2:1:1:2:2
and a homogenization step by ball milling during 1 hour. Kesterite was
prepared with 2.38 g of Cu, 1.23 g of Zn, 2.23 g of Sn, 1.20 g of S and 2.96
g of Se, for each ampoule. Quartz ampoules were filled with the
components homogenised and three cycles were performed alternating
nitrogen and vacuum, finally, the ampoule was sealed containing nitrogen
inside.
Exit of the furnace to avoid impurities of copper, zinc
and/or tin powder dragged by the air flow, in the recovery
of selenium.
3.
Exit of the quartz reactor to avoid loss of selenium
through the assembly.
Elements inside the quartz ampoule reacted for 48-72 h at 500 ºC, in a
tubular furnace or in a muffle.[10] Finally, kesterite mineralized was obtained
and characterized by X-Ray Fluorescence (XRF), Inductively Coupled
Plasma Mass Spectrometry (ICP-MS), Raman Spectroscopy and X-Ray
Diffraction (XRD).
At the exit of the reactor, the gases went to a Dreschel bottle with a solution
of 25% sodium hydroxide to neutralize sulphur dioxide. Temperatures
studied were 700, 730 and 750 ºC and air flows of 20, 40 and 60 mL/min.
Thermal oxidation of kesterite
Kesterite powder was submitted to thermal oxidation for selenium
oxidation and selenium dioxide extraction, in three different pieces of
equipment: a muffle, a furnace of lower capacity and finally, in a tubular
furnace in vertical configuration.
Muffle
Extraction of selenium from kesterite was studied in a muffle of 15 L
capacity. Tests were done with series of 1.0 g of synthetic kesterite in
ceramic crucibles and at temperatures of 400, 500, 600, 700 and 800 ºC.
The residual sample was analysed after each hour of reaction by XRF and
a mass balance was performed. Reaction was considered finished when
more than 95% of selenium present was extracted from the sample.
Figure 5. Experimental set up for thermal oxidation of kesterite in a
vertical tubular furnace.
Small capacity furnace
Selenium was condensed in the zone of the quartz reactor at room
temperature, between glass wools 2 and 3. For the recovery of selenium,
glass wool 1 was extracted to recover metals oxides from the residual
sample, then, glass wool 2 and 3 were removed to recover selenium
condense by washing the inner zone of the reactor with distilled water. The
aqueous solution contains soluble selenium dioxide and insoluble red
selenium that were filtered out, washed with water, dried by suction, and
weighed. The solid selenium was dissolved in HCl and added to the water
solution and analysed by ICP-OES giving the date of total %Se recovery.
After any experiment, the mixture of Zn, Cu, Sn oxides were dissolved in
HCl acid and analysed by ICP-OES (Table 4). The %Se extraction is
calculated from the data of %Se remaining in the sample.
Thermal treatment was also studied in a furnace of 0.5 L of capacity with
a graphite crucible. At the exit of the furnace, a glass hood was placed for
the recovery of selenium and the gas flow was directed to a Dreschel bottle
and a vacuum pump (Figure 4). 1 g of synthetic kesterite powder was
introduced in the graphite crucible of the furnace and heated at
temperatures from 700 to 1000 ºC. The residual sample was analysed
each two hours of reaction. The resulting sample was analysed by XRF
and a mass balance was performed to quantify extraction of selenium.
Reduction of selenium dioxide
The experiments were performed using commercial selenium dioxide to
obtain the optimal processing conditions, which were applied later to the
5
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.202000261
FULL PAPER
treatment of selenium dioxide obtained from thermal oxidation of kesterite
absorbers.
Firstly, Riley reaction was performed, in this process selenium dioxide is
reduced to selenium and the methylene groups adjacent to carbonyl
groups, oxidized to carbonyl.[13] Riley method was studied with two
different carbonyl species: 2-phenylacetophenone and 4-acetylbiphenyl.
As it can be seen in Figure 6, 2-phenylacetophenone is oxidized to 1,2-
diphenylethane-1,2-dione, whereas 4-acetylbiphenyl is oxidized to 4-
biphenylyl (oxo) acetaldehyde (Figure 7).
Validation of the developed process for selenium recovery
Photovoltaic cells of kesterite were provided on Soda Lime Glass (SLG)
substrate. This thin film PV cells were composed of molybdenum, kesterite
2
absorber, cadmium sulphide (CdS), zinc oxide (ZnO ), indium tin oxide
(ITO) and indium (In) metal. The cells were encapsulated at 150-160 ºC
with filler and a front sheet made of plastic.
Kesterite powder was recovered from photovoltaic cells by two main steps:
manual disassembly and chemical treatment. First, the de-capsulation was
achieved by thermal treatment at 90 ºC during 1 hour and manual
disassembly of the plastic and the filler. Then, chemical treatment with
hydrochloric acid 37%, was applied during 30 minutes to separate kesterite
absorber from the PV cell. Treatment with HCl 37% dissolved the front
Figure 6. Riley reaction using 2-phenylacetophenone.
2
layers (CdS, ZnO , ITO and In), molybdenum remained attached to the
In both experiments, selenium dioxide (2.0 g) was dissolved in a mixture
of distilled water and glacial acetic acid (7 mL: 7 mL). Then, reducing agent
SLG and kesterite was released to solution. Kesterite remained insoluble
and therefore was recovered by filtration, washed with distilled water to
neutral pH and dried at 110ºC during 1 h.
(3.6 g) was added, and the resulting solution was heated at 90 ºC and
stirred for 3 hours. The black selenium precipitated was recovery by
filtration and washed with diethyl ether and distilled water, in the case of 2-
phenylacetophenone; nevertheless, the precipitated selenium was
washed with hot water to avoid precipitation of 4-acetylbiphenyl oxidized,
using 4-acetylbiphenyl as reducer.
Kesterite absorber (2.0 g) recovered from PV cells was submitted to
thermal oxidation in a vertical tubular furnace at 750 ºC with an airflow of
20 mL/min. Total reaction time was 3 h 15 min, distributed in 15 minutes
to achieve setpoint, 2 hours of reaction and 1 hour to cool down the reactor
to room temperature, maintaining the airflow.
Selenium condensed in the upper zone of the quartz reactor was
recovered by washing with distilled water and then, reduced to elemental
selenium with ascorbic acid in HCl 0.5 M heating at 70 ºC, during 1.5 h.
Elemental selenium precipitated in solution and it was recovered by
filtration, washed with distilled water to neutral pH, and dried at 110 ºC
during 1 hour. Finally, selenium purity was analysed by XRF and ICP-
OES.
Figure
7.
Riley
reaction
using
4-acetylbiphenyl.
The use of alternative reducing agents, zinc, tin chloride and ascorbic acid
was investigated in order to further increase the yield and efficiency of the
reduction process, as well as to reduce the cost of the recycling process,
and to minimize its environmental impact. Reduction reactions were
studied for 1.0 g of commercial selenium dioxide in 0.5 M hydrochloric acid
and stoichiometric quantity of the studied reducing agents. Ascorbic acid
provided the best results, compared the others reducers, reducing
selenium dioxide to elemental selenium and being oxidized to
dehydroascorbic acid (Figure 9).
Acknowledgements
This research was supported by the H2020 Programme under the
project STARCELL (H2020-NMBP-03-2016-720907). Authors
would like to thank Dr. Florian Oliva and Dr. Susan Schorr for the
help during Raman spectroscopy and X-Ray Diffraction analyses,
respectively. The authors thank Servicio General de Apoyo a la
Investigación−SAI (Universidad de Zaragoza) for technical
assistance.
Keywords: Selenium • Kesterite • Waste recovering • PV cells •
Circular Economy
Figure 8. Reduction of selenium dioxide with ascorbic acid.
[1]
A. Luque, S. Hegedus, Handbook of Photovoltaic Science and
Engineering, John Wiley & Sons, London, 2003.
[
2]
National Renewable Energy Lab. (NREL), End-of-life management.
Solar
Photovoltaic
Panels.
https://www.irena.org/publications/2016/Jun/End-of-life-
management-Solar-Photovoltaic-Panels, 2016 (accesed 20 June
2019).
M. Monteiro, J. Alvarez-Gaitan, J. Bilbao, R. Corkish, in Solar
Panels and Photovoltaic Materials (Ed.: B. Zaidi), Intechopen,
London, 2018.
Ascorbic acid (3 g, 18 mmol) was added to a solution of 200 mL distilled
water and 10 mL of HCl, commercial selenium dioxide (1 g, 9 mmol) was
added and the solution was stirred at 70 ºC during 1.5 h. Selenium reduced
was recovered by filtration, washed to neutral pH and dried at 110 ºC
during 1 hour to be analysed by XRF.
[
[
3]
4]
A. Khatibi, F. Razi, M. Ahmadi Energy Sci Eng. 2019, 7, 305–322.
6
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.202000261
FULL PAPER
[
5]
J. K. Larsen, F. Larsson, T. Törndahl, N. Saini, L. Riekehr, Y. Ren,
A. Biswal, D. Hauschild, L. Weinhardt, C. Heske, C. Platzer-
Björkman, Adv Energy Mater. 2019, 9, 1900439-1900447.
[
6]
World Health Organization, (Trace elements in human nutrition and
health. https://apps.who.int/iris/handle/10665/37931,
accesed 7 July 2019).
F. M. Fordyce, in Essentials of Medical Geology: Revised Edition
Ed.: O. Selinus), Springer Dordrecht, The Netherlands, 2013
C.R.M. Alliance, Critical Raw Materials.
http://criticalrawmaterials.org/critical-raw-materials/, 2018 (accesed
3 March 2019).
1996
(
[
[
7]
8]
(
1
[
[
9]
S. Giraldo, Z. Jehl, M. Placidi, V. Izquierdo-Roca, A. Pérez-
Rodríguez, E. Saucedo, Adv Mater 2019, 31, 1806692-1806710.
10] V. Kheraj, K. K. Patel, S. J. Patel, D. V. Shah, J. Cryst. Growth 2013,
62, 174-177.
3
[
[
11] C.W. Clark, E.M. Elkin, Process for making selenium dioxide,
US2406666A, 1946.
12] E.C. Handwerk, W.T. Monson, Selenium recovery process,
US2948591, 1960.
[
13] H. L. Riley, J. F. Morley, N. A. C. Friend, J J. Chem. Soc. 1932,
875-1883.
1
[
[
14] G. R. Waitkins, C. W. Clark, Chem. Rev. 1945, 36, 235-289.
15] A. M. K. Gustafsson, Chalmers University of Technology
(
Gothenburg, Sweden), 2014.
16] C. Wang, S. Li, H. Wang, J. Fu, J S Afr I Min Met 2016, 116, 593-
00.
17] A. M. K. Gustafsson, M. R. S. Foreman, C. Ekberg, Waste Manag.
014 34, 1775-1782.
[
[
[
[
6
2
18] P. Patnaik in Handbook of inorganic chemicals, McGraw Hill
Professional,New York, NY, 2002.
19] a) D. Fraga, R. Martí, I. Calvet, T. Stoyanova Lyubenova, L. Ladeira
De Oliveira, J. B. Carda, Boletin de la Sociedad Espanola de
Ceramica y Vidrio 2014, 53, 5; b) H. Xie, Y. Sánchez, S. López-
Marino, M. Espíndola-Rodríguez, M. Neuschitzer, D. Sylla, A.
Fairbrother, V. Izquierdo-Roca, A. Pérez-Rodríguez, E. Saucedo,
ACS Appl. Mater. Interfaces. 2014, 6, 12744-12751; c) I. Calvet, E.
Barrachina, R. Martí, D. Fraga, T. Stoyanova Lyubenova, J. B.
Carda, Mater. Lett. 2015, 161, 636-639.
[
[
20] G. G. M. Dimitrievska, H. Xie, A. Carrete, A. Cabot, E. Saucedo, A.
Pérez-Rodríguez, S. Schorr, V. Izquierdo-Roca, J. Alloys Compd.
2015, 628, 464-470.
21] J. Márquez, M. Neuschitzer, M. Dimitrievska, R. Gunder, S. Haass,
M. Werner, Y. E. Romanyuk, S. Schorr, N. M. Pearsall, I. Forbes,
Sol Energ Mat Sol C 2016, 144, 579-585.
[
[
22] A. R. Denton, N. W. Ashcroft, Phys. Rev. 1991, 43, 3161-3164.
23] J. J. Scragg, T. Ericson, T. Kubart, M. Edoff, C. Platzer-Björkman,
Chem. Mater. 2011, 23, 4625-4633.
[
24] S. De Llobet, J.L. Pinilla, R. Molines, I. Suelves Appl. Catal. 2015
165, 457-465.
7
This article is protected by copyright. All rights reserved.