Journal of New Technology and Materials

JNTM
Vol. 06, N°01 (2016)60-72 OEB Univ. Publish. Co.

Quartz sand beneficiation using magnetic and electrostatic

separation to glass industries
Ben Fradj Manel1, Gallala Wissem2 and Abdeljaouad Saadi1
1
Geology Department, Faculty of Sciences, Tunis-El Manar University , Tunisia.
2
Higher Institute Of Fine Arts, University of Sousse, Tunisia.
bf.manel@hotmail.fr

Received date: May 06, 2016; revised date: June 28, 2016; accepted date: June 28, 2016

Abstract

Quartz sand of Fortuna formation was assigned to the Oligo-Miocene. This formation outcrops in Central Tunisia,
particularly in the Ain Bou Morra area. The grain particle size ranges from fine to medium. The morphoscopic analysis
shows that the useful fraction (100-630μm) consists essentially of transparent quartz grains. The mineralogical study of
samples after separation in heavy liquid indicates that they contain a small amount of heavy minerals such as: tourmaline,
zircon and staurotide. The X-ray diffraction analysis of the total rock revealed that quartz is the major mineral constituent
of sand. Chemical analysis shows high content of SiO . Coloring elements (Fe O ) and (TiO ) are slightly elevated. The
2 2 3 2

study aim was to remove impurities from silica sand, in order to upgrade quartz sands and to produce material that has a
higher potential value for industrial manufacturing processes. Several processing physical techniques (attrition, gravity,
magnetic and electrostatic separation) have been developed. The obtained material after treatment was characterized using
Atomic Absorption Spectroscopy (AAS) and Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-AES).
Chemical results through the combined techniques show a significant increase of impurities (such as Fe, Ti, Cr …) and a
significant increase of SiO . The final concentrate achieved 99.99 % SiO , 8 ppm Fe O and 6 ppm TiO was obtained, at
2 2 2 3 2

the optimum operating conditions, from an ore containing about 98.8% SiO , 0.16% Fe O and 0.05% TiO . The treated
2 2 3 2

sand has been found to be a satisfactory material conforms to the requirements of optical glass, crystallaboratory glass and
photovoltaic cells.

Keywords: Attrition, Electrostatic and magnetic separation. Quartz sand.

lower than 0.2%). This kind of raw sand, called extra-

siliceous sand is very abundant in Tunisia, especially in
1. Introduction Oligocene and Miocene outcrops. Different authors
have been studied the characterization of these sand
Silica sand industry in Tunisia has been growing deposits (Griffiths 1987; El Maaoui 1993; Jouirou
rapidly due to increased demand from civil 1981; Louhaichi 1981; Trabelsi 1988; Jamoussi 1991;
engineering, pharmaceutical practices, chemical, Added 2005; Ben Fradj 2010). However, few works
foundries, glass, ceramic, electronics and photovoltaic carried out on the purification and beneficiation
industries. However, the glassmaking is an important process (Aloui, 2010; Gallala et al. 2009; Gallala 2010;
component in the glass production accounting for Gaied & Gallala 2011)
around 65 to 70% wt of total raw material input. Depending on the degree of purification described can
The specifications made on glassmaking sands hold two treatment schedules, which were tested in
(Harben and Kuzvart 1997) are defined essentially by this study.
their chemical composition (SiO , Al O , Fe O and
2 2 3 2 3 • The first schema contains the size classification
TiO ) and by the useful granulometric fraction (0.1 to
2 before and after attrition followed by a dry magnetic
0.6 mm). A high content of SiO which is more than
2 separation.
98% is combined with low impurities (%Fe O must be 2 3
 Quartz sand beneficiation using magnetic …. JNTM(2016) Ben Fradj Manel et al.

A wet screening on 1.7 to 1mm before attrition to (governorate of Kairouan, central Tunisia.).
remove large and mid-large quartz coated tablets The site is located east of the sub-meridian structure
which can cause mechanical problems within the Boudabouss. The area is covered by the geological
attrition cell. Djebibina map.
A classification after attrition ranging from 0.1 to 0.63 In this region the series show outcrop of Tertiary age
mm was used, in order to recover the useful fraction deposits .These deposits are affected by sub-meridian
and remove the fine fraction below 0.1 mm. The latter direction of wrinkles. The sands studied belong to the
has contents in harmful elements significantly higher upper Fortuna formation of Aquitanian age.
than the coarse fractions. The series are composed of three lithological units
This scheme allows the preparation of purified sands from the bottom to the top (Figure 2):
with iron contents of about 300 to 400 ppm and a 68% - Lower unit: it shows 30 m thick sand-clay alternations
weight yield for Aquitania sand from Ain Bou Morra. covered by a large mass of misclassified sand that
• The second scheme includes more gravity exceeds 70 m thick. Indeed, this basal series is in the
separation, between the attrition operation and high form of consolidated sandstone rock benches with a
intensity magnetic separation wet and still followed by general rate of granular decreasing sequences stratified
electrostatic separation. This scheme was applied on sandstones.
three composite samples of sand Ain Bou Morra - Middle unit: with 200 m thick, it is
helped prepare sands much more refined, with iron characterized by medium particle size sand with cross-
content in the order of 9ppm and a weight of 76% stratification.
yield. - Upper unit: with 345 m thick, it is
characterized fine sand very ranked well.
The sands are organized in decreasing size sequences
2. Geological Setting with coarse sand quartz at the bottom and finer sand at
the top of each sequence.
The study area is located in the region of Ain Bou The sequences are intercalated by centimetric clay
Morra. It is 20 Km from west of Sbikha village layers

Figure 1 : Location of the study area on an extract of the geological map 1/500 000 of Tunisia

61
 Quartz sand beneficiation using magnetic …. JNTM(2016) Ben Fradj Manel et al.

Location of the studied samples.

MI – s : Langhien – Serravallien, Clay interbedded of sands (MAHMOUD). Mb –I : Burdigalien supérieur – Langhien, Sandy limestone
with bioplaste (AIN GRAB). Ma –b : aquitanien – Burdigalien a- coarse sandstone with quartz dragees (EL HAOURIA higher),
b- Alternating sandstone and clay (EL HAOURIA lower). O : Oligocène Clay interbedded with sands and limestones

Figure 2 : Location of the study site at the geological map Djebibina of 1:50 000.

Figure 3 : Geological section in the region of Ain Bou Morra

3. Material and methods surface. The samples were mixed to prepare a fairly
representative sample of each faces. The obtained
sample has undergone several quartages to obtain
Sand samples were collected by making grooves similar samples with equal amount for analytical
perpendicular to the outcrops and after etching the processing.

62
 Quartz sand beneficiation using magnetic …. JNTM(2016) Ben Fradj Manel et al.

The representative ore sample is obtained from Ain electrostatic separation tests as well .These two tests
Bou Morra Province, Tunisia. For the determination were carried out at the laboratory of Liege University.
of granulometric distribution, standard sieves from the
series “AFNOR” were used.
Major elements were analyzed by Atomic Absorption 4. Results and discussion
Spectrophotometer (AAS) “Perkin Elmer” 3300. In
addition for monitoring minor and trace elements, the 4.1 First method of treatment
Inductively Coupled Plasma Atomic Emission
Spectrometer (ICP-AES) (ULTIMA-C) was The first method includes three stages: the
employed. The mineralogical analysis was carried out classification (screening) before and after attrition
using stereomicroscopy. petrographic microscope and followed by a dry magnetic separation.
X-ray diffraction “X’ Pert Pro MPD PANalytical”
using Cu Kα radiation operating at 40KV and 20mA 4.2 Raw Sand characterization
were used.
The heavy minerals separation was carried out by the 4.2.1 Mineralogical analysis
standard tetrabromoethane technique (specific gravity
The observation under a binocular microscope shows
2.96). In this case, the attrition was done using
transparent grains with irregular shape. About 99% of
Wemco cell (sand scrubber).
the grains are formed by small quartz grains.
This process consisted of conditioning the sample at
The raw samples under the binocular microscope,
70% solids during 15min. Size fractions from +0.1 to
X-ray diffraction analysis: were conducted XRD
0.63 mm were used for gravity separation by shaking
reveals the dominance of quartz phase (3.34 Å) for the
table. Dry magnetic separator (15000 Gauss), wet high
analysis of medium sand samples.
intensity magnetic separator (20,000 Gauss) and
However XRD analysis for the fraction > 2µm
electrostatic separator (25 to 30KV) from Carpco were
identified the presence of kaolin and illite clay.
respectively used for magnetic separation and

Figure 4 : X ray diffractogram clay treated with ethylene glycol

63
 Quartz sand beneficiation using magnetic …. JNTM(2016) Ben Fradj Manel et al.

Figure 5 : X ray diffractogram of clay heated at 550°C

Figure 6 : X ray diffractogram powder of raw sample 1

4.2.2 Chemical and grain size analysis These elements which are concentrated in the clay
fractions. Major impurities, qualified as penalizing
Results of chemical analyses (Table 1) reveal that agent in the glass industry, mainly Fe O and TiO
2 3 2

quality of silica sand is reasonable but some high range respectively from 0.12 to 0.17 wt% and from
impurities make uncertainty to produce high grade 0.033 to 0.08 Wt%. These contents
glass or photovoltaic cells. Major elements, mainly are outside the specifications for high grade glass. In
(SiO ), ranges from 98.69 % to 99.55wt% which gives
2 order to obtain satisfactory product, it is primordial to
an extra siliceous character for the sand (Table 1). upgrade the raw materials.
Although contents, in some samples, of Al O 2 3

(0.25 to 0.47 wt %), Na O (0.012–0.017 wt%), K O

2 2

(0.11–0.17 wt%) are not tolerable, but it is possible to

remove or decrease .

64
 Quartz sand beneficiation using magnetic …. JNTM(2016) Ben Fradj Manel et al.

Table 1 : Chemical analysis of raw sand samples

Sample SiO2 Al2O3 CaO Fe2O3 K2 O MgO Na2O TiO2 Cr Cu

% % % % % % % % (ppm) (ppm)
Sample 1 99.39 0.25 0.07 0.12 0.11 0.010 0.012 0.033 9 8
Sample 2 99.17 0.29 0.09 0.15 0.13 0.015 0.017 0.04 10 9
Sample 3 98.69 0.47 0.11 0.17 0.17 0.07 0.016 0.08 9 9

Figure 7 : X ray diffractogram powder of raw sample 2

Figure 8 : X ray diffractogram powder of raw sample 3

65
 Quartz sand beneficiation using magnetic …. JNTM(2016) Ben Fradj Manel et al.

Plate 1 : photos of Mineralogical identification of the conductive portion of Fortuna sand Ain Bou Morra

66
 Quartz sand beneficiation using magnetic …. JNTM(2016) Ben Fradj Manel et al.

Plate 2 : Mineralogical identification of the dense fraction of Fortuna sand from Ain Bou Morra

67
 Quartz sand beneficiation using magnetic …. JNTM(2016) Ben Fradj Manel et al.

4.2.3 Physical treatments physical treatment is necessary comprising the

following successive stages:
The raw fraction (-0.63+0.1mm) is suitable for the The first schema contains the size classification before
manufacture of flat glass and other types of white glass. and after attrition followed by a dry magnetic
However, to make high-grade silica sand an additional separation attrition and electrostatic separation. The
results of the final product are (Table 2):
Table 2 : results of chemical analysis of the final product after treatment by attrition and dry magnetic separation

SiO2 Al2O3 CaO Fe2O3 K2 O MgO Na2O TiO2 Cr Cu

Sample
% % % % % % % % (ppm) (ppm)
Sample 1
99.51 0.14 0.03 0.031 0.11 0.0072 0.013 0.01 nd nd
Sample 2
99.39 0.12 0.05 0.029 0.10 0.0061 0.012 0.012 nd nd
Sample 3
99.36 0.11 0.026 0.022 0.09 0.0073 0.015 0.01 nd nd
nd: not detected

For the three studied samples, the purified fraction Table 4 : distribution of weight and the% of Fe2O3 according to the
after attrition and dry magnetic separation is the fraction not conductive voltage variation (%)
available size fraction between 0.1et 0.63mm. % Weight of
VOLTAGE
Purified samples show silica content of 99.36%. non-conductive % Fe2O3
KV
The contents of the most harmful impurities fraction
decreased after treatment and may have an iron 10 40 0,03
content between 0.02 and 0.031%, a content of 0.11 to 15 71 0,025
0.14% for alumina, a content of 0.01 % of titanium; 20 89 0,0028
chromium being not detected. 28 98 0,0001
30 79 0,0025
The second scheme includes more gravity separation,
between the attrition operation and wet high intensity Table 5 : distribution of weight and the% of Fe2O3 according to the
magnetic separation (WHIMS) and followed by fraction not conductive voltage variation (%)
electrostatic separation. % Weight of
To reach the requirements of high-quality sand VOLTAGE non-conductive
industry, the sand should be further purified with silica KV fraction % Fe2O3
grains devoid of other minerals or any contaminants. 10 37 0,029
The final product must include a total contaminant 15 66 0,023
level <10 ppm (Outotec, 2007). In order to reach this 20 81 0,004
specification the high voltage electrostatic separation 28 94 0,0001
was applied. 30 80 0,002
Practically, the separator with different voltage value is
used. The results are given in table 3.

Table 3 : Variation of% weight and the% of Fe2O3 of the fraction

non-conductive fraction
% Weight of
VOLTAGE
non-conductive % Fe2O3
KV
fraction
10 42 0.028
15 72 0.019
20 87 0.003
28 99 0.0009
30 82 0.007
Figure 9. Curve of variation of the weight% of the non-conductive
portion in function of the voltage (kV): case of the average sample
Am1.

68
 Quartz sand beneficiation using magnetic …. JNTM(2016) Ben Fradj Manel et al.

Figure 11. Curve of variation of the weight% of the non-conductive

portion in function of the voltage (kV): case of the average sample
AM2.
Figure 10. Variation curve of the Fe2O3 content in function of the
voltage (kV) If the average sample AM1.
The best yield weight is obtained with 28 kilovolts. If
the voltage exceed this threshold the result becomes
less efficient.
Indeed, by applying a voltage higher than 28 kilovolts
a double failure is present: firstly the manifestation of a
very intense electrical discharge and secondly the
substantial decrease of the weight of non-conductive
fraction yield (Example AM1 sample the% weight
decreases from 99 to 82%). The voltage of 28 kilovolts
is the ideal voltage for efficient separation while
maintaining the other parameters constantThe results
of chemical analyses of the non-conductive fraction are
represented in table 6.
Figure 12. Variation curve of the Fe2O3 content in function of the
voltage (kV) If the average sample AM2.

Table 6 : Results of chemical analyses of the final product after attrition treatment, gravity separation, wet high intensity magnetic separation and
electrostatic separation.
SiO 2 Al O
2 3 CaO Fe O
2 3 KO2 MgO Na O2 TiO 2 Cr Cu
Sample
% % % % % % % % (ppm) (ppm)
Sample 1
99.99 0.008 0.005 0.0009 0.009 0.002 0.003 0.001 nd nd
Sample 2
99.97 0.01 0.009 0.0010 0.009 0.004 0.003 0.002 nd nd
Sample 3
99.94 0.03 0.011 0.0013 0.01 0.003 0.004 0.004 nd nd
nd: not detected

It is shown that the depletion of iron oxide (0.0009% This type of testing has also been undertaken with
Fe2O3) and titanium oxide (0.001 %TiO2) in the non only attrition of sands. The beneficiation of the sand is
conductive fractions (Table 6) was successfully less effective.
conducted. These induce that gravity and wet magnetic separation
This proves that the adopted treatment approach and electrostatic separation, were needed to make an
decrease the potential of contamination for the final efficient recovery process.
product of quartz sand. Therefore, the obtained The processing steps are illustrated in summarized
product could be used in high technology industries flow-sheet.
such as the manufacture of silicon for photovoltaic
cells manufacture, optical fibers and electronic chip.

69
 Quartz sand beneficiation using magnetic …. JNTM(2016) Ben Fradj Manel et al.

Figure 13: Flowsheet of the two presented schemes

70
 Quartz sand beneficiation using magnetic …. JNTM(2016) Ben Fradj Manel et al.

Acknowledgments

The authors would like to express their sincere

thanks to Stoyan Gaydardzhiev, associate Professor, at
the University of Liege for access to laboratory and
collaboration in microscopic analysis. We also
acknowledge David Bastin for his assistance in the
physical and chemical analysis.

References
Figure 4 Curve of variation of the weight% of the non-conductive
portion in function of the voltage (kV): case of the average sample
- Added M., 2005. Impact de la qualité des
Am3 granulats sur les caractéristiques physico-
mécaniques des mortiers et des bétons. Thèse
Doc. Univ. Tunis El Manar. Faculté des Sciences
de Tunis.175p.
- Aloui T., 2010. Etude sédimentologique et
intérêt économique des sables d’âge barrémien
(formation Sidi Aich) Tunisie Centro-
méridionale. Thèse Doc. Univ. Tunis II,
Tunisie. 287p.
- Ben Fradj M., 2010.Etude minéralogique,
chimique et minéralurgique des sables de Menzel
Mhiri et de Ain Bou Morra. Mémoire de
Mastère. Univ. Tunis El Manar. Faculté des
Sciences de Tunis. 131p.
Figure 5. Variation curve of the Fe2O3 content in function of the - Ben Haj Ali, M., Jedoui, Y., Ben Salem, H.,
voltage (kV) If the average sample AM3. Memmi, L., 1985.Carte géologique de la Tunisie
: 1/500 000. Office National des Mines, Service
Géologique.
- Gaied M.S., 1991. Géologie des matériaux utiles
Conclusion de la région d’El Gnater (Tunisie Centrale).
Mémoire de DEA. Univ. Tunis El Manar.
Faculté des Sciences de Tunis. 118p,
Mineralogical and chemical studies carried out on - Gaied M. E, Gallala W., 2011. Beneficiation of
silica sand samples from Ain Bou Morra area shows feldspar ore for application in the ceramic
that the main impurities are iron, titanium and industry: Influence of composition on the
feldspar minerals. Most of the impurities are physical characteristics. Arabian Journal of
distributed in the finer fraction (flat glass, optical glass, Chemistry, V.8, 2, p186–190.
glass white, half white glass, glass wool and glass brick) - Gallala, W. (2010). Les sables quartzo-
and also for the chemical industry (silicate feldspathiques de la Tunisie centro-méridionale:
manufacturing soda). sédimentologie, minéralogie, minéralurgie et
applications industrielles, PhD Thesis, Faculty of
While under the second scheme, prepared sands can Sciences Sfax, Tunisia.
be used even for the manufacture of crystal glass as - Gallala W, Gaied M. E., Montacer M., 2009.
raw material for silicon production which is the main Concentration of Potassium Feldspar from Low-
element used for the manufacture of photovoltaic
grade Sidi Aïch Sand in Tunisia for Industrial
solar cells and electronic chips.
Applications. Silicates industriels, no 5-6, p. 125-
130.
- Griffits J., 1987. Silica. Is the choice crystal clear.
Industrial Minerals, n° 235. Pp 25-43.

71
 Quartz sand beneficiation using magnetic …. JNTM(2016) Ben Fradj Manel et al.

- El Maaoui M., 1993. Chimie pratique du verre. gouvernorat de Kairouan (Tunisie) application à
Rapport, CTMCCV. la recherche et à la cartographie des matériaux
- Harben, Peter W., KuŽvart, M., 1997. Industrial utiles. Thèse Doc. Univ. Franche-Comté. Faculté
minerals: a global geology. Metal Bulletin. des sciences et des techniques. 146p.
- Jamoussi F., 1991. Etude géologique et - Ounis M.S., 2004. Contribution a l’étude du
géotechnique des substances minérales utiles de sable Oligocène de Oueslatia application aux
la région de Gafsa. Thèse de spécialité. Faculté domaines de l’industrie. Diplôme de mastère en
des Sciences de Tunis. 298p. géologie, Faculté des sciences de Tunis. 67p.
- Jouirou M., 1981. Etude géologique et - Trabelsi, H., 1989. Les matériaux utiles de la
géochimique des sédiments de la région d’El Kef région de Feriana (Tunisie centrale) (cadre
(NW de la Tunisie). Thèse Doc. Univ. de géologique et potentialités). Thèse de doctorat,
Bordeaux.146p. Nice.
- Louhaichi M.L., 1981. Etude géologique et
géotechnique de la partie nord-occidentale du .

72