Available online at www.sciencedirect.com
ScienceDirect
Availableonline
onlineatatwww.sciencedirect.com
www.sciencedirect.com
Available
Energy Procedia 00 (2017) 000–000
ScienceDirect
ScienceDirect
www.elsevier.com/locate/procedia
Energy Procedia
Procedia 00
136(2017)
(2017)000–000
53–59
Energy
www.elsevier.com/locate/procedia
4th International Conference on Energy and Environment Research, ICEER 2017, 17-20 July
2017, Porto, Portugal
Recovery
silicon
carbide
from on
waste
silicon
by using
Theof15th
International
Symposium
District
Heating slurry
and Cooling
flotation
Assessing the feasibility of using the heat demand-outdoor
Lia,*,Wei-Sheng
Chenaheat demand forecast
temperature functionHsun-Chi
for a long-term
district
a
Department of Resources Engineering, National Cheng Kung University, No.1, University Road, Tainan City, Taiwan, Republic of China
I. Andrića,b,c*, A. Pinaa, P. Ferrãoa, J. Fournierb., B. Lacarrièrec, O. Le Correc
a
IN+ Center for Innovation, Technology and Policy Research - Instituto Superior Técnico, Av. Rovisco Pais 1, 1049-001 Lisbon, Portugal
b
Veolia Recherche & Innovation, 291 Avenue Dreyfous Daniel, 78520 Limay, France
c
Département Systèmes Énergétiques et Environnement - IMT Atlantique, 4 rue Alfred Kastler, 44300 Nantes, France
Abstract
The amount of waste silicon slurry increases as the wafer production raises. The slurry is generally disposed using incineration or
land-filling. Separating high-purity SiC from waste silicon slurry can reduce costs for enterprises and assist in waste reuse and
recycling. In this study, flotation was applied to separate SiC and Si from waste silicon slurry through hydrophilicity and
Abstract
hydrophobicity
of the particle surface. By controlling the concentration of hydrofluoric acid and the oxidation reduction potential
of the two stages flotation, the SiC can be separated from Si. The optimal condition of first stage flotation is 0.8 mol/L of HF at District
addressed
in theinliterature
as stage.
one ofUnder
the most
solutions
for decreasing
the
400
mV, heating
while 0.6networks
mol/L ofare
HFcommonly
and -300 mV
was applied
the second
theseeffective
conditions,
approximately
52.8% of
greenhouse
gas emissions
building
SiC
was recovered
with the from
gradethe
of 98.1
%. sector. These systems require high investments which are returned through the heat
sales. Due to the changed climate conditions and building renovation policies, heat demand in the future could decrease,
©prolonging
2017 The the
Authors.
Published
byperiod.
Elsevier Ltd.
investment
return
Peer-review
underofresponsibility
of assess
the scientific
committee
of thethe
4thheat
International
on Energyfunction
and Environment
The main scope
this paper is to
the feasibility
of using
demand – Conference
outdoor temperature
for heat demand
forecast. The district of Alvalade, located in Lisbon (Portugal), was used as a case study. The district is consisted of 665
Research.
buildings that vary in both construction period and typology. Three weather scenarios (low, medium, high) and three district
Keywords:
Flotation;
recycling;
carbide;
silicon; waste
silicon slurry
renovation
scenarios
were silicon
developed
(shallow,
intermediate,
deep). To estimate the error, obtained heat demand values were
compared with results from a dynamic heat demand model, previously developed and validated by the authors.
The results showed that when only weather change is considered, the margin of error could be acceptable for some applications
error in annual demand was lower than 20% for all weather scenarios considered). However, after introducing renovation
1.(the
Introduction
scenarios, the error value increased up to 59.5% (depending on the weather and renovation scenarios combination considered).
The
value
of slope
coefficient
increased
average within
range
of 3.8% upraised
to 8%since
per decade,
that slicing
corresponds
to the
The
amount
of waste
silicon
slurryon
increased
as the the
wafer
production
the wafer
primarily
decrease in thecutting
numberofofsilicon
heatingingots
hours into
of 22-139h
during This
the heating
season
(depending
the saw,
combination
weatherand
and
accompanied
thin wafers.
process
requires
a steel on
wire
grindingof media,
renovation
considered).
On the other
function
intercept
increasedEnvironmental
for 7.8-12.7% per
decade (depending
on the
coolant
andscenarios
thus caused
the impurities
in thehand,
slurry.
According
to Taiwan
Protection
Administration
coupled scenarios). The values suggested could be used to modify the function parameters for the scenarios considered, and
(EPA), approximately 20,000 tons of waste silicon slurry was produced annually in the solar industry in Taiwan. The
improve the accuracy of heat demand estimations.
© 2017 The Authors. Published by Elsevier Ltd.
Peer-review under responsibility of the Scientific Committee of The 15th International Symposium on District Heating and
Cooling.
* Corresponding author. Tel.: +886-6-2757575 ext. 62828; fax: +886-6-3840960.
E-mail address: n48021034@mail.ncku.edu.tw
Keywords: Heat demand; Forecast; Climate change
1876-6102 © 2017 The Authors. Published by Elsevier Ltd.
Peer-review under responsibility of the scientific committee of the 4th International Conference on Energy and Environment
Research.
1876-6102 © 2017 The Authors. Published by Elsevier Ltd.
Peer-review under responsibility of the Scientific Committee of The 15th International Symposium on District Heating and Cooling.
1876-6102 © 2017 The Authors. Published by Elsevier Ltd.
Peer-review under responsibility of the scientific committee of the 4th International Conference on Energy and Environment
Research.
10.1016/j.egypro.2017.10.281
Li et al.Procedia
/ Energy00
Procedia
136 (2017) 53–59
Li andHsun-Chi
Chen / Energy
(2017) 000–000
54
waste silicon slurry was generally disposed using incineration or land-filling and therefore leads to environmental
damage and wasting substantial resources. The mainly components of the slurry are silicon carbide (SiC) and silicon
(Si) mixed with a small amount of iron and alcohol compounds like PEG, PG and DEG (T. Surek, (2005)). It is
expected that SiC and Si can be recycled from the waste slurry and reused in the related industry such as wafer
production and steel casting to reduce the costs and particularly protect the environment.
Waste silicon slurry can be separated using physical, chemical methods or flotation. Physical separation mainly
includes centrifugation, electrical field and magnetic separations. Wang et al. (2009) used heavy-fluid high-gravity
centrifugation with temperature treatment and directional solidification to separate SiC from Si recovering 86% of SiC
with 87% of grade. Lin and his team (2010, 2010) combined centrifugation with phase-transfer separation to recover
Si from silicon slurry. The phase-transfer separation improved the grade of recovered Si from 90.8% to 99.1%. By
controlling the surface potential through centrifugation, Lin et al. (2013) recycled 90.8% of SiC with 95.2% of grade.
Wu and Chen (2009) separated SiC and Si through different particle velocity under different electrical field. The grades
of the recycled SiC and Si were 90% and 92.4%, respectively. Tsai et al. (2011, 2013) investigated the efficiency of
electrical field separation under different electric field strength, baffle plate. The grade of SiC and Si could be improved
to 95.2% and 92.9%, respectively. For magnetic separation, Sergii et al. (2014) experimented high-gradient magnetic
separation combined with hydrocyclone and sedimentation. With different specific weight, 95% grade SiC was
recycled from the slurry. Nishijima et al. (2003) removed iron from waste silicon slurry by employing superconduction magnetic separation and recycled SiC by centrifugation. The recovery efficiency of SiC was 80%. The
main chemical separation method is liquid-liquid extraction. Wei et al. (2015) separated SiC and Si via different
hydrophobic characteristics of particle surfaces, and established a silicon regression model from the results of the
research.
The influences of activator, collector and frother on flotation were investigated separately (Iskra, 1997, Lin et al.,
2002, Sahoo et al., 2016). Shibata (2006) conducted flotation separation using various cationic surfactants to separate
SiC and Si. Instead of directly separating the waste silicon slurry, Shibata recovered SiC through flotation after silicon
oxidation into SiO2 and obtained a 99.7% grade SiC with a recovery efficiency of 96.7%. As regards previous research,
the hydrofluoric acid (HF) dissolved the SiO2 layer on the surface of particles in waste silicon slurry (Eq. 1) and made
SiC a hydrophilic polar sinking mineral (Larsen and Kleiv, 2016). In addition, the floatability of silicon in the slurry
increased since the Si-F bonding on the surface of Si particles (Ernsberger, 1960, Guo et al., 2005, Lin et al., 2008).
SiO2 + 6HF → H2 (SiF6 ) + 2H2 O
(1)
This study focus on the flotation separation of SiC and Si rather than SiO2. In order to raise the recovery efficiency
and grade of recovered SiC, a surface activator was applied. The concentration of surface activator was investigated
to achieve the optimal condition leading to high recovery efficiency and grade. The oxidation reduction potential
(ORP) of the system was also controlled to verify the optimal potential for silicon removal. High grade of SiC can be
recovered as concentrate by controlling the oxidation-reduction potential and the hydrophobicity of Si particles.
2. Experimental
2.1. Composition and properties
The sample for this study is the powder dried from the waste silicon slurry produced from the wafer fabrication.
The SiC grade of the waste silicon slurry was determined by acid leaching, the waste silicon slurry was leached using
hydrofluoric, nitric, sulfuric, and hydrochloric acids, after which the sample was filtered and subjected to hightemperature drying. An X-ray diffractometer (XRD, Bruker AXS-D8A) was used to analyze the crystalline phase of
the sample; the particle size distribution was determined using a laser particle size distribution analyzer (LS, Beckman
Coulter LS230); and the particle morphology was observed using a scanning electron microscope (SEM, Hitachi
S3000-N).
Hsun-Chi
et al. /Procedia
Energy Procedia
136
(2017) 53–59
Li and
Chen /LiEnergy
00 (2017)
000–000
55
2.2. Flotation separation
This study investigated the application of floatation separation to separate SiC and Si in waste silicon slurry. The
flotation equipment was capable of separating nonferrous metal, noble metal, and nonmetallic minerals and materials.
The most suitable particle size is 80 to 150 um (H.K.Lin et al. (2002)). The flotation in this study was arranged as a
two-stage process running continuously but individually. The equipment (Denver sub-A Cell) is a single-trough
flotation unit featuring a 3L pulp capacity, a rotation speed of 300–3300 rpm. During the process, the flotation
equipment generates negative pressure by centrifugation and blade rotation to mix the pulp and chemical agents and
produces refined foam. Floating minerals were scraped off after they had attached to the foam, during which the scraper
rotation speed was adjusted, air was injected, and surface characteristics of particles were controlled.
2.2.1. Surface activator
In this study, hydrofluoric acid was used as surface activator to modify the characteristics of the surface of particles.
The surface characteristics was modified to change the Si particles from hydrophilicity into hydrophobicity and thus
the slurry can be separated through flotation.
2.2.2. Concentration of hydrofluoric acid
Since the surface characteristic of Si particles had been changed into hydrophobic by hydrofluoric acid, the
concentration was also investigated. The concentration of HF varied from 0.2 to 1.0 mol/L in the flotation process to
find out the optimal concentration of HF with the highest recovery efficiency and grade of SiC. Considered the twostage flotation, the concentration of HF in each stage was assessed individually.
2.2.3. Oxidation reduction potential
Floatation separation was employed to primarily investigate changes in ORP in pulp. The ORP was controlled
within -500 to -100 mV to obtain the highest recovery efficiency with high grade SiC. A flotation reagent was added
to control the potential of the slurry and improve flotation recovery efficiency. The particle surface becomes either
hydrophilic or hydrophobic when its electric charge changes under a specific ORP of the pulp and reagent
concentration. The ORP was calculated using the Nernst Equation (Eq. 2).
Ea = E0a +
where:
𝐸𝐸 𝑎𝑎
𝐸𝐸𝑎𝑎0
z
[ox]
[red]
0.05916
n
[ox]
log [red]
(2)
: reduction potential
: standard reduction potential
: number of moles of electrons transferred in the cell reaction
: activity of the oxidized form
: activity of the reduced form
The waste silicon slurry in the flotation separation had a mass of 300 g and a solid-to-liquid ratio of 1:5. hydrofluoric
acid was used as the interface activator. The pulp was separated by the hydrophilicity and hydrophobicity of SiC and
Si. The separation was divided into first stage and second stage flotation to determine the difference in concentration
of HF and ORP in addition to the SiC recovery grade and recovery efficiency.
3. Results and discussion
3.1. Waste slurry characterization
Table 1 summarizes the waste silicon slurry composition. The XRD spectrum revealed that the sample mainly
contains SiC and Si (Fig. 1). The SEM shows that there are two kinds of particle size, which are SiC and Si (Fig. 2).
The LS analysis of the slurry is presented in Fig. 3, the larger particle size is 7~20 µm and the small particle size is
0.7~3 µm. The SiC particle size is 9.24 µm and Si particle size is 1.0 µm according to Tsai (2013). Therefore, in Fig. 2
Hsun-Chi
Li et al.Procedia
/ Energy00
Procedia
136 (2017) 53–59
Li and
Chen / Energy
(2017) 000–000
56
that the larger particles are SiC and the smaller ones are Si. The mean and median particle size of the waste silicon
slurry was approximately 3.976 µm and 1.708µm.
Table 1. Composition of waste silicon slurry.
wt ％
SiC
Si
Fe
K
Ca
51.2
45.6
2.4
0.3
0.5
Fig. 1. XRD pattern of waste silicon slurry.
Fig. 2. SEM of waste silicon slurry.
Fig. 3. Particle size distribution of waste silicon slurry.
3.2. Flotation results
3.1.1. Effect of surface activator
Table 2. Effect of the activator to the recovery efficiency and grade (%).
Activator
No activator
Hydrofluoric acid
Tailings
Concentrate
Grade
Recovery
Grade
Recovery
50.6
42.6
9.4
84.2
51.3
97.3
90.6
15.8
Hydrofluoric acid was used as surface activator in this study since particle surfaces of SiC and Si contain a SiO 2
layer. As shown in Table 2, the grade of the SiC in the tailings and concentrate is almost the same without surface
activator. Using HF, the grade of SiC in the concentrate raised to 97.3%. Not only the SiO2 was dissolved by the HF,
Hsun-Chi
et al. /Procedia
Energy Procedia
136
(2017) 53–59
Li and
Chen /LiEnergy
00 (2017)
000–000
57
but the Si-H bonding on the surface of Si particles was replaced by Si-F bonding and caused the hydrophobic of Si
particles. On the contrary, the SiC particles became hydrophilic thanks to the smaller contact angle once the SiO2 layer
was removed.Although the recovery efficiency of the concentrate in the flotation with HF is very low, the surface
activator is necessary for SiC recycling due to the high grade of recovered SiC. The concentration of HF and the ORP
of the pulp were further investigated in order to achieve the optimal conditions leading to high recovery efficiency and
high grade.
3.1.2. Effect of hydrofluoric acid concentration
The second part of the experiment was to evaluate the effect of the HF concentration on the two-stages flotation
process operating continuously but individually. During the separation of the waste silicon slurry, 300 gram of waste
slurry was applied at a solid-liquid ratio of 1:5. In the first stage of the flotation, the concentration of HF varied from
0.4 to 1.0 mol/L. Fig. 4 shows that using 0.4 mol/L of HF, the recovered SiC with 53.3% of grade and the efficiency
of 53.4% means deficiency of concentration of HF. With 1.0 mol/L of HF, the recovery efficiency and grade of
recovered SiC decreased. The optimal concentration of HF was determined as 0.8 mol/L, corresponding to 80.1% of
recovery concentrate with the SiC grade of 74.6%.
The results from the second stage flotation are shown in Fig. 5. A similar situation of stoichiometry overdose was
observed for 0.8 mol/L HF. 0.6 mol/L of HF was considered the optimal concentration in this second stage. With the
two-stage flotation considering the concentration of HF, the grade of recovered SiC is 98.2% and recovery efficiency
is 76.6%. The pulp in the first stage floatation required more HF to provide fluoride ions to form hydrophobic surface
of the Si particles since the higher Si content of the slurry. After first separation, the lower Si content reduces the
concentration of HF from 0.8 to 0.6 mol/L. However, the concentration of HF cannot be increased continuously. The
stoichiometry overdose changes the hydrophilicity of SiC particles into hydrophobicity, thus resulting poor separation
efficiency.
Fig. 4. Concentration effect of HF to SiC grade and recovery
efficiency (1st stage).
Fig. 5. Concentration effect of HF to SiC grade and recovery
efficiency (2nd stage).
3.1.3. Effect of oxidation reduction potential
The ORP experiment was also investigated, arranging as a two-stage flotation and investigated continuously but
individually. The ORP was calculated using the Nernst Equation. The potential was controlled from -500 to -100 mV
in the first stage of ORP flotation. Fig. 6 shows the recovery efficiency and grade of SiC after the first stage. With 400 mV of the pulp potential, the grade of SiC was 72.5% with recovery efficiency 78.2%. The potential was controlled
a little positive, -350 to -100 mV, thanks to the lower Si content. The results of the second stage of ORP flotation are
showed in Fig. 7. Considering the optimal ORP at -300 mV for the second stage, the optimal grade of the recovered
SiC was 97.5% with 84.3% of recovery efficiency of the two-stage flotation considering the ORP.
In the absence of a SiO2 layer on the Si and SiC surfaces, Si–F bonds were formed on the Si particle surfaces,
releasing hydrogen ions and electrons and producing more of the reduced form in the slurry, which was described by
Hsun-Chi Li et al. / Energy Procedia 136 (2017) 53–59
Li and Chen / Energy Procedia 00 (2017) 000–000
58
a negative ORP. The SiC particles had OH− on their surfaces and thus did not affect the changes in ORP.
Fig. 6. ORP effect on SiC grade and recovery efficiency (1st stage).
Fig. 7. ORP effect on SiC grade and recovery efficiency (2nd stage).
3.3. Concentrate analysis
The concentrate of the flotation experiment was analyzed through by SEM and LS to see the differences of the
slurry powder after the flotation process. Fig. 8 is the SEM diagram of the concentrate. The particle size in Fig. 8
increased obviously after the flotation when compared to the untreated waste slurry (Fig. 2). In addition, the particle
size distribution is shown in Fig. 9. The particle size was centered at 10 μm, which means the finer particles (<5μm)
of silicon were removed through flotation. The mean and median particle size of the slurry was summarized in Table.
3. Both the mean and median particle size raise indicate the enrichment and purification of SiC. The mean and median
particles size of the concentrate was 9.294 and 9.695 μm.
From this study, flotation was applied to separate and recover the SiC in the waste silicon slurry. In the first stage
of flotation, 0.8 mol/L of HF was used as surface activator under the ORP condition of -400 mV. The second stage
of flotation was controlled at -300 mV with 0.6 mol/L of HF. By controlling the experiment under the optimal condition
of two-stage, the overall grade of recovered SiC is 98.1% whereas the overall recovery efficiency of SiC is 52.8%.
Table 3. Effect of the flotation process to the particle size.
(μm)
mean
median
Before
After
3.967
9.294
1.708
9.695
Fig. 8. The SEM of concentrate.
Fig. 9. Particle size distribution after flotation.
Hsun-Chi
et al. /Procedia
Energy Procedia
136
(2017) 53–59
Li and
Chen /LiEnergy
00 (2017)
000–000
59
4. Conclusion
This study revealed that flotation separation can be used to separate waste silicon slurry by controlling the oxidation
reduction potential of the pulp to change the hydrophilicity and hydrophobicity of SiC and Si, using HF as the interface
activator. According to the aforementioned results, the findings of this study are described as follows:
1. Comparison of the particle size analysis from before and after by flotation, mean and median particle size is
3.976 µm and 1.708 μm before flotation and mean and median particle size is 9.294 µm and 9.695μm after
flotation. Using the flotation method, the silicon carbide in the waste silicon slurry can be recovered.
2. The optimal concentration of HF in the two-stage flotation process is 0.8 mol/L for first stage and 0.6 mol/L
for the second stage, leading to 76.6% of SiC recovery efficiency and 98.2% of grade.
3. The optimal oxidation reduction potential in the two-stage flotation process is -400 mV for the first stage and
-300 mV for the second stage, which allowed obtaining 84.3% of SiC recovery efficiency and 97.5% of grade.
4. The overall results of this study, operation at the optimal condition of HF concentration and ORP process,
were 52.8% of SiC recovery efficiency with 98.1% of grade.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
Thomas Surek. (2005),” Crystal growth and materials research in photovoltaics: progress and challenges.” Journal of Crystal growth 275
(2005): 292-304.
T. Y. Wang, Y. C. Lin, C. Y. Tai, C. C. Fei, M. Y. Tseng and C. W. Lan. (2009),” Recovery of silicon from kerf loss slurry waste for
photovoltaic applications.” Progress in photovoltaics: research and applications 17 (2009): 155-163.
Yen-Chih Lin and Clifford Y. Tai. (2010),” Recovery of silicon powder from kerfs loss slurry using phase-transfer separation method.”
Separation and purification technology 74 (2010): 170-177.
Yen-Chin Lin, Teng-Yu Wang, Chung-Wen Lan and Clifford Y. Tai. (2010),” Recovery of silicon powder from kerf loss slurry by
centrifugation.” Powder Technology 200 (2010): 216-223.
Suning Lin, kai Huang and Hongmin Zhu. (2013),” Recovery of powder from silicon wiresawing slurries by tuning the particle potential
combined with centrifugation.” Separation and purification technology 118 (2013): 448-454.
Yung-Fu Wu and Yuan-Ming Chen. (2009),” Separation of silicon and silicon carbide using an electrical field.” Separation and purification
technology 68 (2009): 70-74.
Tzu-Hsuan Tsai. (2011),” Modified sedimentation system for improving of silicon and silicon carbide in recycling of sawing waste.”
Separation and purification technology 78 (2011): 16-20.
Tzu-Hsuan Tsai, Yu-Pei Shih and Yung-Fu Wu. (2013),” Recycling silicon wire-saw slurries: Separation of silicon and silicon carbide in a
ramp settling tank under an applied electrical field.” Journal of the Air & Waste Management Association 63.5 (2013): 521-527.
Sergii A. Sergiienko, Boris V. Pogorelov and Vladimir B. Daniliuk. (2014),” Silicon and Silicon carbide powders recycling technology from
wire-saw cutting waste in slicing process of silicon ingots.” Separation and purification technology 133 (2014): 16-21.
Shigehiro Nishijima, Yoshinobu Izumi, Shin-Ichi Takeda, Hiroki Suemoto, Atsushi Nakahira and Shin-Ichi Horie. (2003),” Recycling of
abrasives from wasted slurry by superconducting magnetic separation.” IEEE Transactions on applied superconductivity 13.2 (2003): 15961599.
Xiu-qin Wei, Chuan-qiang Yin, Yue-peng Wan and Lang Zhou. (2015),” Effect of wet oxidation on recovery of silicon from wire saw slurry
by liquid-liquid extraction.” Separation and purification technology 149 (2015): 457-461.
J. Iskra. (1997),” Flotation properties of silicon carbide I: Flotation of silicon carbide with anionic and cationic collector.” Ceramics
International 23 (1997): 337-342.
H. K. Lin, J. L. Oleson, J. T. Hollow and D. E. Walsh. (2002),” Characterization and flotation of gold in carbon fine at the Fort Knox mine
Alaska.” Minerals and Metallurgical Processing 19.1 (2002): 21-24.
Hrushikesh Sahoo, Swagat S. Rath, Danda S. Rao, Barada K. Mishra and Bisweswar Das. (2016),” Role of silica and alumina content in the
flotation of iron ores.” International Journal of Mineral processing 148 (2016): 83-91.
Junji Shibata, Norihiro Murayama and Kengo Nagae. (2006),” Flotation separation of SiC from wastes in the silicon wafer slicing
process.”Kagaku Kogaku Ronbunshu 32.1 (2006): 93-98.
E.Larsen, R. A. Kleiv. (2016),” Flotation of quartz from quartz-feldspar mixtures by the HF method.” Minerals Engineering 98 (2016): 4951.
F. M. Ernsberger. (1960),” Structural effects in the chemical reactivity of silica and silicates.” Journal of physics and chemistry of solids 13
(1960): 347-351.
Zhiguang Guo, Feng Zhou, Jingcheng Hao and Weimin Liu. (2005),” Stable biomimetic super-hydrophobic engineering materials.” Journal
of the American Chemical Society 127 (2005): 15670-15671.
H. K. Lin, D. E. Walsh, S. H. Sonderland, C. Bissue and A. Debrah. (2008),” Floatability of metallic iron fines from comminution circuits
and their effect on flotation of a sulfide ore.” Minerals and Metallurgical Processing 25.4 (2008): 206-210.