Journal of Minerals & Materials Characterization & Engineering, Vol. 11, No.2 pp.183-192, 201 2
jmmce.org Printed in the USA. All rights reserved
183
Chemical and Mineralo gica l Characterization of Ghanaian Found ry Slags
A. Andrews1*, E. Gikunoo1, L. Ofosu-Mensah1, H. Tofah1, S. Bansah2
1Department of Materials Engineering, Kwame Nkrumah University of Science and Technology,
Kumasi-Ghana
2Department of Geological Engineering, Kwame Nkrumah University of Science and Technology,
Kumasi-Ghana
*Corresponding Author: anthonydrews@gmail.com
ABSTRACT
The application of ferrous slags as aggregate, road bed materials, soil fertility conditioner and
as clinker materials depend on their chemical and mineralogical characteristics. In order to
assess the possibilities of using ferrous slags generated in Ghana, selected slag samples from
some foundries have been characterized. A total of six bulk samples were collected from three
different foundry shops to determine their chemical and mineralogical composition. The slag
samples showed varied chemical and mineralogical characteristics. Differences are as a result of
the differences in furnace charge, operation temperatures, grades of refractory furnace lining
and melting additives. Calcium oxide (CaO) contents observed in all investigated samples were
in combined form (i.e. not free CaO). The low amounts of calcium oxide in the slag samples limit
their applications in the area of cement/aggregate production. Even though slags were found to
be alkaline, for these slags to effectively neutralize the acidity of soil and improve its fertility,
components such as CaO, SiO2, MgO and FeO in the slag must be improved. However, these
slags could be used as road bed materials.
Keywords: Cupola slag; Induction furnace slag; Characterization; Recycling
1. INTRODUCTION
In the Ghana foundry industry, only minute fraction of the wastes generated is recovered and
reused. Large fraction of the waste generated are often disposed off on the factor y floor. Foundry
wastes generated includes waste foundry sand, slag, dust and sludge. Usually, these wastes are
184 A. Andrews, E. Gikunoo, L. Ofosu-Mensah Vol.11, No.2
hazardous to the environment [1-3] . For instance, slag containing sulfurous leachate can result in
sulfur odor and discoloration of water in poor drainage conditions [4]. Additionally, slag
containing high levels of nickel can cause corrosion of iron and steel in the presence of moisture
[5]. Nevertheless, slags containing negligible amount of sulfur and nickel are not ecologically
harmful and therefore could be used [6]. Therefore, there is the need to properly handle foundry
slag to alleviate any potential adverse effects on the environment.
Slag is a waste material (by-product) generated in purifying metals, their casting and alloying.
Ferrous slags are generated from the production and castings of iron and steel. There ar e d if fe rent
types of ferrous slags, each one named from the process of which they are generated: blast
furnace s lag (B F) also ca lled iron s lag, basi c ox ygen furnac e slag (BO F), el ectric ar c furna ce sl ag
(EAF) and ladle furnace slag (LF) also called refining slag. Slag types investigated in this work
are cupola furnace (CF) slag and induction furnace (IF) slags [7].
Current interviews and data collection suggest that more than 50,000 tons of ferrous slags are
produced in Ghana annually. This figure will continue to rise as small and medium scale
foundries spring up. Additionally, until quality scraps, energy efficient furnaces, quality
refractory furnace linings and quality foundry coke are used, the amount of slag generated would
increase as these parameters directl y impact on slag generation and its quality [8-10]. Currently,
slag produced is disposed off on the foundry site floor. However, there is a great demand for
aggregate in the Civil Engineering industry in Ghana, especially in the construction of road and
production of Portland cement. In other countries, such as Western Europe and in the U.S.A,
blast furnace slag is widely known to serve as raw material for cement production as well as
concrete manufacturing unit due to the presence of pozzolonic activity [6,11,12]. Electric arc
furnace (EAF) slag cannot be used for such purposes due to its high iron content. EAF has
however, found increased use as recycled aggregate in road construction [12]. Steel slag having
high content of CaO but low levels of phosphorus and sulfur are reused as fluxing material in
steel making process to replace limestone [7]. Steel slags are also being used in areas such as
road construction, fertilizers and soil conditioners, and for metal recoveries. Utilizations of these
slags were made possible by first knowing their physico-chemical and mineralogical properties.
A proper characterization of foundry slag could effectively serve as basis for identifying possible
utilization areas in Ghana. There is no literature available on the chemical and mineralogical
characteristics of cupola and induction furnace slags generated in Ghana. To date, slags from
these furnaces are not being used and are deposited on the site. These slags are air-cooled and are
therefore crystalline [13].
In the present study an attempt has been made to determine the chemical and mineralogical
composition of cupola and induction furnace slags generated by some small and medium scale
foundries in Ghana. Knowledge of these is important in order to know the possible utilization
areas of these by-products.
Vol.11, No.2 CHEMICAL AND MINERALOGI CAL CHARACTERIZATION 185
2. MATERIALS AND METHODS
2.1 Material s
Induction furnace (IF) slag samples were obtained from Western Steel and Forging Limited
located i n the Accra-Tema m et ropo li s. Cupola furnace (CF) sla g sam pl es were obtained from two
small scal e fou nd ries (Horoos and Abudia foundries) located in the Kumasi metropolis. Sampling
was carried ou t b y inserting a sample rod into a slag pan as the slag was tapped. The rod was then
withdrawn from the pan and allowed to cool in air. Two of such slag samples were taken after
each steel production batch. Samples were crushed with a jaw crusher followed by milling to
obtain powdered samples. The powdered samples were sieved with a 250 µm size.
Representative samples of the undersize fraction were used for further analyses. Samples A and B
are repres entative CF slags from Horoos and Abudia foundries, respectively. Sample C is IF slag
from Western Steel and Forging Limited.
2.2 Method
The chemical compositions of the slags were determined using X-ray fluorescence spectroscop y
(XRF- Spectro X-LAB 2000). The samples were sieved and the undersize from a 180 µm mesh
grid obtained. 4g of ea ch repres en tat iv e sam pl es were taken and mix ed with equal amount of wax
for 3 minutes in a homogeniser. Th e mixtures were compressed i nto circul ar shapes for th e XRF
analysis. XRF analyses were carried out in duplicate and the average results taken.
Mineralogical composition was studied at room temperature (25oC) using X-ray diffractometer
(Siemens D5000), operating at 40kV and 40mA, using Cu Kα radiation. S am pl es were s cann ed at
15o – 70o (2θ) and step size of 0.02o. Samples were then milled to pass 10 µm mesh grid before
scanning. The pH of the slag samples was determined using a pH meter. Ganulometric analysis
(mesh grid) was carried out to determine size distribution in the slag samples investigated.
3. RESULTS AND DISCUSSION
3.1 Physical Characteristics of Slags
All samples were air cooled. Samples had blue black appearance. A typical image of the ferrous
slag is shown in Figure 1. Investigated samples were alkaline with sample A having pH = 8.06,
sample B pH = 8.20 and sample C pH = 7.14.
Figure 2 shows the ganulometric analysis of the slag samples. The differences in % passing for
the three ferrous slag samples varied slightly with sample A and sample B (copula slags) having
higher % passing than sample C. The samples investigated had particle sizes lower than 425 µm
with only 2% passing 75 µm sieves. The grain s ize anal ysis i s wid ely used in the clas sifi cation of
186 A. Andrews, E. Gikunoo, L. Ofosu-Mensah Vol.11, No.2
slag. The data obtained from grain size distribution curves is used in the design of filters for earth
dams and to determine suitability of slag for road construction [14].
Fig. 1: Typical image of ferrous slag.
Fig. 2: Size distribution curves of slag samples.
3.2 Chemical Composition
The summary of the chemical composition of the slag samples investigated is shown in Table 1.
The major constituents in the sl ag samp les, determi ned by XRF, were Na2O, MgO, Al 2O3, SiO2,
P2O5, K2O, CaO, TiO2, MnO and Fe2O3. The minor metal constituents are present as traces,
with a con centrati on les s than 16400 ppm. Trace elements determined include V, Cr, Ba, Co, Ni,
Cu, Zn, Ga, As, Rb, Sr, Y, Zr, Nb, Mo, Sn, Sb, Cs, La, Ce, Hf, Ta, Pb, Bi and U. There were
variations in the chemical composition of the three samples investigated. Sample B had higher
percent of major oxides (Al2O3, SiO2, Na2O, MgO, K2O, CaO) than samples A and C. It was
Vol.11, No.2 CHEMICAL AND MINERALOGI CAL CHARACTERIZATION 187
observed that, even for cupola furnace slag samples, there were strong variations in chemical
composition. Sample A, for instance, had strong variations in Na2O (6.07%), Al2O3 (12.76%),
and Fe2O3 (8.59%) contents compared to sample B which had 15.90% Na2O, 18.16% Al2O3,
and 3.67% Fe2O3. The difference could be attributed to one or a combination of the following
reasons: (1) contamination from refractory furnace linings, (2) quality and quantity of scraps
charged into the furnace, (3) type and quantity of fluxing agent used and (4) type and quantit y of
coke used. Small and medium scale foundries in Ghana use different furnace charge ratios
(coke:scraps:flux) without carrying out proper m aterial balances. This results in reduced melting
efficiency and therefore increasing generation of slag. High amount of Fe2O3 was recorded for
induction furnace slag (sample C). This is due to the slag removal technique which is by
scooping or tilting of the furnace whilst in cupola furnaces, slag is allowed to flow through the
slag spout. Table 1 also shows the loss on ignition (LOI) of the investigated samples. Sample A
recorded the highest LOI of 18.20% followed by sample C (16.80%) and sample B (3.80%),
respectively. Once again, the difference could be attributed to varied amount of furnace charge.
Table 1: Chemical compositions of investigated slag samples.
Composition (%)
A
B
C
Na
2
O
6.07
15.90
0.48
MgO
1.31
2.00
1.68
Al
2
O
3
12.76
18.16
8.45
SiO
2
45.82
31.51
48.54
P
2
O
3
0.09
0.09
0.06
SO
3
0.59
0.45
0.37
Cl
0.03
0.00
0.08
K
2
O
0.27
0.41
0.32
CaO
4.19
5.97
3.91
TiO
2
0.55
0.47
2.82
MnO
1.79
0.55
7.62
Fe
2
O
3
8.59
3.67
25.89
LOI
18.20
3.80
16.80
3.3 Mineralogical Composition
XRD analysis was performed on the as-received samples. Figure 3 shows the XRD pattern for the
investigated slags. The diffraction patterns were different for the three ferrous samples even
though all slags are chemically similar in that their major components are CaO, SiO2, Al 2O3 and
MgO. All slag types were heterogeneous materials consisting of a mixture of complex crystalline
phases. The analysis of the diffraction lines revealed the presence of the following major
constituent phases: quartz (SiO2), kyanite (Al 2SiO5), maghemite (γ-Fe2O3), calcium iron oxide
188 A. Andrews, E. Gikunoo, L. Ofosu-Mensah Vol.11, No.2
(CaFe4O7), and kanoite (Mn, Mg)2(Si2O6). Al l slags inv est i gated did not contain free CaO, M gO
and elemental iron. Iron present was in the form o f an oxide or i n solid solut ion. An y other minor
constituent phases were very difficult to assign due to the complexity of the diffractograms, the
low intensity of their diffraction lines as well as strong background signals.
Fig. 3: XRD patterns of slag samples.
3.4 Possible Industrial Utilization
The amount of ferrous slags produced in foundry industries depends on the type of scraps, coke
quality, and furnace efficiency among others [2-4,15]. Large amount of slag is produced when
light scraps (rusted scraps) are used. Usually steel scraps produced the least amount of slag [16].
At Western Steel and Forging Limited, about 15 tons of slag is produced for every 100 tons of
steel ingot produced. Small scale foundries produce about 5 tons of slag for ever y 30 tons of steel
produced. It is interesting to note from this data that both small and medium scale foundries
produce roughly the same amount of slag. As mentioned earlier, this is because of the complexit y
of various parameters that affect slag generation in the foundry operations. Unpublished data
indicate that there are over 50 small and medium scale foundries scattered across Ghana. From
our survey results, it is estimated that the amount of ferrous slag produced yearly is about
500,000 tons. The amount of slag generation in this industry would increase as foundrymen
continue to charge rusted scraps, low quality furnace linings, and poor energy efficiency
furnaces.
A
B
Vol.11, No.2 CHEMICAL AND MINERALOGI CAL CHARACTERIZATION 189
Metallurgical slags have been used as raw materials for reclamation, cement production, road
construction, floor tiles and fertilizers production, or soil conditioning [17-20]. The presence of
major oxides like CaO, SiO 2, FeO and Al 2O3 make metallurgical slag useful as clinker material
in cement productions [21]. The clinker reduces the amount of energy used in the calcinations
process b ecaus e of the p resence of free C aO. It also reduces the amount of C O 2 released i nto the
atmosphere [22].
The min eralo gical an al ysis i ndicat ed that the s lag sam ples i nvest igated h ad n o free Ca O. Fo r sla g
to be useful as raw material in cement production, it must meet certain requirements. In
accordance with the European Standard (EN-197-1), slag to be used for cement production must
have a minimum basicity ratio of 1 (i.e. (CaO+MgO):SiO2 ≥ 1). In practice, for technological
reasons, the Al2O3:Fe2O3 ratio (must be < 1), or so called hydraulic properties, expressed by the
relatio ns hi p (C aO+MgO+Al2O3)/SiO2 (must be > 1) should also be known [7]. It is interesting to
know that, even though the chemical composition of slags has complicated structures, they are
usually expressed as if these compounds existed separately within the material. These oxide
ratios have been calculated and tabulated in Table 2. In addition, slag to be ut ilized in the cement
industry should contain the least possible contents of elements such as sodium, potassium,
chlorine, chromium, as well as heavy metals and alkalis [7,20-22]. From Table 2, it is observed
that the slag samples investigated had basicity ratios to be less than 1 which do not meet the
requirem ent for u se in clin ker material s. Onl y sample C had Al2O3:Fe2O3 rat io to be less than 1,
however, the basicity ratio was also less than 1 and therefore does not also satisfy the
requirement. The hydraulic properties of the investigated slags are also not satisfactory since the
oxide ratios are less than 1.
Table 2: Ratios of chemical composition of investigated slags.
Slags
A
2.15
1.49
0.40
0.12
B
1.44
4.95
0.83
0.25
C
1.41
0.33
0.29
0.14
The pres ence o f free C aO an d MgO i n metal lur gical slags is the reason wh y the y are not us ed fo r
road construction. It is believed that the presence of these oxides results in expansion of slags
(about 10 vol. %) under the influence of humidity hydrate [7,23,24]. When the presence of free
CaO accounts for more than 1 % increase in volume, another adverse property of steel slag,
namely, the appearance of white powder in the form of sediment begins to form. Free CaO from
the leachate bounds with water, creating calcium hydroxide, Ca(OH) , which, when exposed to
atmospheric conditions, reacts with carbon dioxide, creating calcium carbonate (C aCO3) [23,24].
This settles down in the form of white powder and may cause obstructions in the drainage
systems and water retention. Those obstructions are particularly dangerous in the case of
190 A. Andrews, E. Gikunoo, L. Ofosu-Mensah Vol.11, No.2
freezing, which renders large damage to pavement structures [7]. The investigated slags do not
have free CaO and MgO and therefore can be used for road bed construction.
Soil is generally acidic. Alkaline slags are used to amend soil acidity. Soil fertility is improved by
the presence of SiO2, MgO and FeO in slag [25,26]. The Ca and P contents in alkaline slags
make it a potential liming agent to increase the precipitation and sorption of metals such as Cu
and a potential fertilizer promoting plant growth and improving physico-chemical properties of
the soil. Slags investigated are alkaline and can therefore be used to amend soil pH. It must
however be mentioned that for effective results, the CaO, MgO, FeO and SiO2 contents in the
slag be improved.
4. CONCLUSION
Chemi cal anal yses of sla g sam ples i nvest igated showed varied compositio n as follows: CaO 3.91
- 5.97%, MgO 1.31 - 2.00%, Al 2O3 8.45 - 18.16%, SiO 2 31.51 - 48.54%, Na 2O 0.48 - 15.90%,
Fe2O3 3.67 - 25.89%, MnO 0.55 - 7.62% and K2O 0.27 - 0.41%. Ratios of oxide composition
suggest that these slags cannot be used as clinker material for the production of Portland cement.
From the chemical and mineralogical characterization results, it can be concluded that, cupola
furnace slag and induction furnace slag generated in Ghana can only be useful as road bed
material. However, CaO content must still be improved by addition of lime or proper slag mixing
ratios for effective results. Slag samples are also alkaline and can therefore be used as soil
fertility improver with slight modification in CaO content.
REFERENCES
[1] Ghosh, S., Das, M., Chakrabarti, S., Ghatak, S. 2002, Development of ceramic tiles from
common clay and blast furnace slag. Ceram Int. 28, 393–400.
[2] Osipov, V.A., Timofeeva, Z.G., Kurbatskii, M.N., Mironova, P.V., Ocheretnyuk, F.F.
1990, Blast furnace granulated slag in the production of building ceramics. Glass Ceram.
47, 228–230.
[3] Dana, K., Das, S.K., 2004, Partial substitution of feldspar by blast furnace slag in triaxial
porcelain: Phase and microstructural evolution. J. Euro. Ceram. Soc. 24, 3833–3839.
[4] Karamanova, E., Avdeev, G., Karamanov, A., 2011, Ceramics from blast furnace slag,
kaolin and quartz, J. Euro. Ceram. Soc. 31, 989–998.
[5] Nistor, L., Cioroi, M., 2007, Recycling possibilities of metallurgical slag, The Annals of
Dunarea DeJos, University of Galati, Fascicle IX, Metallurgy and Materials Science (1),
78-82.
[6] Fidancevska, E., Vassilev, V., Hristova-Vasileva, T., Milosevski, M., 2009, On a
possibility for application of industrial wastes of metallurgical slag and TV-glass, J. Univ.
Chem. Technol. & Met. 44, 2, 189-196.
Vol.11, No.2 CHEMICAL AND MINERALOGI CAL CHARACTERIZATION 191
[7] Navarro, C., Diaz, M., Villa-Gacia, M. A., 2010, Physico-chemi cal charac terizatio n of steel
slag. Study of its behavior under simulated environmental conditions. Environ. Sci.
Technol. 44, 5383-5388.
[8] Adygezalov, S.A., Dvortsov, S.P., Shemonaeva, T.N., Strakhov, V.M., Surovtseva, I.V.,
2011, Industrial production of coke briquets for cupola smelting of iron, Coke & Chemistry,
54 (1), 19-22.
[9] Diez, M.A., Alv ar ez, R., Barriocanal, C., 2002, Coal for metallurgical coke production:
Prediction of coke quality and future requirements for coke making. Int. J. Coal Geol., Vol.
50, 289-412.
[10] Haley, D.G., 2004, The efficiency of coke producing quality iron from lower quality
coke, www.asi-alloys.com/pdf/GeoHaleyCokePaper.pdf.
[11] Sarkar, R., Singh, N., Das, S.K., 2010, Utilization of steel melting electric arc furnace slag
for development of vitreous ceramic tiles, Bull. Mater. Sci., 33 (3), 293–298.
[12] Skuza, Z., Kolmasiak, C., Prusak, R., 2009, Possibilities for the utilization of metallurgical
slag in the conditions of polish economy, METABK, 48(2), 125-128.
[13] Barišić, I., Dimter, S., Netinger, I., 2010, Possibilities of application of slag in road
construction, Technica l Ga zet t e, 17 (4), 523-528.
[14] Sofilić, T., Rastovčan-Mioč, A., Ćosić, M., Merle1, V., Mioč, B., Sofilić, U., 2011, EAF
steel slag application possibilities in Croatia asphalt mixture
production: http://www.aidic.it/CISAP4/webpapers/16Sofilic.pdf.
[15] Arkhipov, N.A., Gorokhov, V.L., 2003, New approaches to processing slags, Metallurgist,
47, (7–8).
[16] Tsakiridis, P.E., Papadimitriou, G.D., Tsivilis, S., Koroneos, C., 2008, Utilization of steel
slag for Portland cement clinker production, J. Hazardous Mater., 152, 805–811.
[17] Alanyali, H., Col, M., Yılmaz, M., Karagoz, S., 2006, Application of magnetic separation
to steelmaking slags for reclamation, Waste Management, 26, 1133–1139.
[18] Asi, I.M., Qasrawi, H.Y., Shalabi, F.I., 2007, Use of s teel sl ag ag greg ate i n as ph alt con crete
mixes, Can. J. Civ. Eng. 34, 902–911.
[19] Badiee, H., Maghsoudipour, A., Dehkordi, B.R., 2008, Use of Iranian steel slag for
production of ceramic floor tiles, Advances in Applied Ceramics, 107 (2), 111-115.
[20] Qiang, W., Peiyu, Y., 2010, Hydration properties of basic oxygen furnace steel slag,
Construction and Building Materials, 24, 1134–1140.
[21] Wang, G., 2010, Determination of the expansion force of coarse steel slag aggregate,
Construction and Building Materials, 24, 1961–1966.
[22] Shi, C., 2004, Steel slag Its production, processing, characteristics, and cementitious
properties, J. Mater. Civil Eng., 231-236.
[23] Le wis, W.D., 2011, Properties and uses of iron and steel slags.
http://www.nationaslagassoc.org.
[24] Song, L.K.M.R., Emery, S., 2001, Preliminary development of slag as a stabilized material
for labor intensive construction of roads. Workshop on first international conference on
employment creation on development, Univ. Witwatersrand, South Africa.
192 A. Andrews, E. Gikunoo, L. Ofosu-Mensah Vol.11, No.2
[25] Alva, A.K., Sumner, M.E., 1990, Amelioration of acid soil infertility by phosphogypsum,
Plant and Soil, 128 (2), 127-129.
[26] Carvalho-Pupatto, J.G., Bull, L.T., Crusciol, C.A.C., 2004, Soil chemical attributes, root
growth and rice yield according to slag application, Pesquisa Agropecuaria Brasileria, 39
(12), 1213-1218.