Professional Documents
Culture Documents
Role of Minor
Elements in Cement
Manufacture and Use
By Javed I. Bhatty
Contents Page
INDEX 1
ABSTRACT 3
KEYWORDS 3
INTRODUCTION 4
DEFINITIONS 7
Major Elements 7
Lesser Elements 9
Minor Elements 9
Trace Elements 9
Contents Page
TRANSITION ELEMENTS 34
Yttrium 35
Titanium 35
Zirconium 37
Vanadium 37
Niobium 38
Tantalum 38
Chromium 38
Molybdenum 40
Tungsten 40
Manganese 40
Cobalt 41
Nickel 41
Copper 42
Silver 42
Zinc 42
Cadmium 43
Mercury 44
CONCLUSIONS 46
ACKNOWLEDGEMENTS 46
APPENDIX 47
ABSTRACT
In this review, the effects of almost all the stable minor and trace elements on the
production and performance of Portland cement have been reported. Emphasis has
been given to elements that occur in natural and by product materials used for cement
manufacturing. The elements for which detailed information has been obtained are dealt
with in an order based on the periodic classification of the elements. The volatilities of
the elements have also been discussed where ever necessary. Elements reviewed
include: hydrogen, sodium, potassium, lithium, rubidium, cesium, barium, beryllium,
strontium, magnesium, boron, gallium, indium, thallium, carbon, germanium, tin, lead,
nitrogen, phosphorous, arsenic, antimony, bismuth, oxygen, sulfur, selenium, tellurium,
fluorine, chlorine, bromine, iodine, helium, neon, argon, krypton, xenon, yttrium, titanium,
zirconium, vanadium, niobium, tantalum, chromium, molybdenum, tungsten, manganese,
cobalt, nickel, copper, silver, zinc, cadmium, mercury, and the lanthanides.
KEYWORDS
INTRODUCTION
The purpose of this review is to collect pertinent information on the behavior of minor
and trace elements on the manufacture and use of cement. Attempts have been made to
identify gaps, if any, in the information thus far available and suggest work for further
investigations.
The effects of minor elements on the production of clinker and performance of cement
are summarized in the appendix.
There has been an increasing move toward using a variety of alternative materials in
cement manufacturing, with the multiple aims of reducing by-product accumulation to
address environmental problems and to achieve technical advantages during clinker
processing without sacrificing the quality of cement.
Chloride by-products, and wastes from the soda ash industry, when mixed with fly ash
and limestone, are reported to have produced low temperature clinkers (1200°C) with
comparable compressive strength (Patel, 1989). Phosphogypsum has been used as a
source of lime in kiln feed. Thought the clinker attained a different microstructure, the
cement compared favorably with the conventional type (Toit, 1989). In separate studies,
spent clays from lubricating oil refining, have also been tested as raw feed components
for clinker production (Midlam, 1985).
Sewage sludge as a partial kiln fuel was reported by Obrist (1987). Heavy metals in the
sludge were permanently withdrawn from the biosphere with little toxic emissions.
Organic pollutants were reliably destroyed without leaving any toxic by-products. The
only exception may be mercury which must be controlled adequately.
Ostrovlyanchik et al. (1986) reported that the use of fly ash from coal power plants as
raw material, in place of argillaceous material, was effective for improving the wet
process kiln output with savings in the fuel consumption.
Bhatty et al. (1985) produced an ASTM C150 type I cement from the copper-nickel and
taconite tailings used as a partial substitute for cement raw feed. The resulting cement
had a microstructure and strength properties comparable to that produced with
conventional feed.
Weatherhead and Blumenthal (1992), of the Scrap Tire Management Council, concluded
from a recent field study that tires can be used successfully as an alternative fuel in
cement kilns. The fuel cost is significantly lowered, and the production rate is enhanced
without adversely affecting the quality of cement. No significant change in the
environment quality, due to emissions, was noted (1).
Use of scrap tires and lubricating oil, as alternative fuel in cement manufacturing, has
also been reported in a number of tests burns in Australia and Canada (McGrath, 1993;
and Heron, 1993). The tests suggested substantial reduction in manufacturing costs,
improvement in waste minimization advances in resource recovery, extended life of
existing landfills, and better environmental control.
(1)
A distinction between the terms “tire-derived fuel” and “whole tires” may be made here.
The tire derived fuel (TDF) uses shredded tires in combination with other conventional fuels
(coal, coke, oil, gas, etc.) usually in the burner end, tire chips are also fed to calciners,
whereas the whole tires are feed end of a precalciner or a preheater kiln or into the calcining
zone of a long kiln.
De Zorzi (1988) reported Italian experience on the use of municipal solid waste as a
partial fuel in cement manufacturing. The chemical and physical characteristics of clinker
and cement were comparable to those produced conventionally. No significant change in
organic or inorganic pollutants, such as dioxins, furans, SO x, and Nox were detected in
the stack emissions. There were no material handling problems, but storage of the solid
waste, especially the refuse derived fuel (RDF), was expensive, because it needed to be
contained in totally enclosed compartments for technical and environmental reasons.
Huhta (1990) surveyed some twenty North America cement plants operating with wastes
as supplement fuel, and noted that the predominant waste being used was waste oil
followed by solvent derived fuel and scrap auto tires; wood chips and fluid coke were
also mentioned. However, it is projected that the tire derived fuel (TDF) will become the
most advantageous fuel in the near future, because of its availability and easy handling.
Kelly (1992), and Mantus et al. (1992) reported that the use of wastes as supplemental
fuel in well designed and properly operated kilns results in metal emissions too negligible
to cause any adverse health effects. It was also demonstrated that the cements and kiln
dusts thus produced were not substantially different from those conventionally produced.
The effect of wastes on the emissions of organic compounds and metals from kilns were
also studied by Von Seebach et al. earlier in 1990. It was reported that a virtually
complete destruction and removal of hazardous organic compounds occurs in the kiln. A
Destruction and Removal Efficiency (DRE) of hazardous compounds was recorded at
99.9996%. DREs in access of 99.97 are routinely achieved.
Siemering, Parsons, and Lochbrunner (1991) have also reported on their experiences of
burning wastes as kiln fuel and have reported both technical and economical
advantages with minimal adverse environmental impact.
In a recent article, Hansen (1993) has strongly advocated the use of solid wastes in
cement manufacturing, emphasizing potential environmental and political advantages. It
was suggested that using wastes as fuel has two-fold environmental benefits; it would
not only avoid fossil fuel extraction and transportation, but would also minimize
emissions that would have occurred by disposal of these wastes through treatment or by
landfilling. Politically, the liability of landfills and demands for waste minimization in an
environmental sensitive society, such as the United States, can be substantially reduced
by waste utilization in cement making.
Gossman (1988) pointed out some of the risks and liabilities associated with the use of
hazardous waste-derived fuels, and proposed certain analytical means to minimize
liabilities in order to achieve full economical and quality control advantages.
Although there are opportunities to beneficially use wastes in cement production, their
total substitution in the industry is still in the experimental stages, One recommendation
has been to limit the use of waste to 5% by weight of the raw feed (Vogel et al., 1987).
Huhta (1990), and Von Seebach et al. (1990) have reported a potential of 20-30% or
even more for waste as a fuel replacement in cement kilns; some plants have already
used 50-100% replacement. Nonetheless, the level of application and degree of success
largely depends upon the waste composition in terms of the type and concentration of
minor or trace elements.
In summary, under favorable practical conditions, wastes can have the following
combined benefits:
DEFINITIONS
Major Elements
According to Rompps Chemie-Lexikon (1987), the elements that are more abundantly
present (>5%) in cement clinker are the major elements. These are calcium (Ca), silicon
(Si), aluminum (Al), iron (Fe), and oxygen (O). Carbon (C), and nitrogen (N), because of
their abundance in the raw material and the earth atmosphere respectively, can also be
regarded as major elements.
In clinker and cement analyses, Ca, Si, Al and Fe are expressed as the oxide form (CaO,
SiO2, Al2O3 and Fe2O3). However, they eventually exist as more complex compounds.
The approximate formulae of these compounds, also known as clinker phases, are
tricalcium silicate 3CaO.SiO2 or C3S (2), dicalcium silicate 2CaO.SiO2 or C2S, tricalcium
aluminate 3CaO.Al2O3 or C3A, and tetracalcium aluminoferrite 4CaO.Al2O3.Fe2O3 or
C4AF. Since the role of major elements in cement manufacturing has been fairly well
understood, only a brief summary on the presence of major compounds in cement is
given here.
A typical composition of ASTM Type I cement, the most commonly used cement in
general construction is normally 58% of C 3S, 15% C2S, 8% C3A, 10% C4AF, and 5% of
gypsum. Other ASTM cement types are Type II, III, IV, and V, which vary in composition
and are used where special properties are required. Typical composition and physical
properties of various cement types are given in Table 1 (adapted from CTL, 1993;
Mindess and Young, 1981).
Type III cement is a high heat of hydration cement with high C3S content and a finer
particle distribution and is used where rapid hardening is required for early strength
development. Type IV is a low heat of hydration, slow setting cement because of low C 3S
and high C2S contents. It is intended for mass concrete in order to avoid thermal
cracking, but is now rarely produced. Since the strength development of type IV cement
is low, Type II cement, which can be specified as a moderate heat of hydration cement,
is generally recommended due to its higher strength and market availability. For even
lower heat of hydration, Type II cement with fly ash is used. Type V cement is also a low
heat of hydration cement because of low C 3S and low C3A contents; it is normally used
when high sulfate resistance is required. Type II is primarily used as a moderate sulfate
resistant cement.
The knowledge of the compound composition can reasonably be used to predict the
properties of cement. One of the known methods for calculating compound composition
from the oxide analysis are the Bogue formulae (1955). Although a number of
sophisticated techniques are now available for Bogue calculations, the simplest Bogue
formulation that has been found suitable for most applications is given in the ASTM C150
specifications.
(3)
Alites and belites are never pure forms of C3S and C2S respectively. Due to the geological
source of the raw materials, alite and belites will always have small quantities of impurities or
trace elements.
Lesser Elements
Four lesser elements, i.e., sodium (Na), potassium (K), magnesium (Mg), and sulfur (S),
which appear in virtually all commercial clinkers at 1-5% concentration, are represented
in chemical analyses as oxide forms: Na2O, K2O, MgO, and SO3. Rompps Chemie-
Lexikon (1987) has termed these elements as the secondary elements. In cement
chemist’s notation Na2O = N, K2O = K, MgO = M, and SO3=S.
Minor Elements
According to Miller (1976) and Gartner (1980), elements other than the major and the
lesser constituents (i.e. Ca, Si, Al, Fe, O, Na, K, Mg, S) may be considered as minor
elements with regards to cement manufacturing. The concentration levels of minor
elements in the clinker are almost always less than 1% and are generally categorized on
the basis of the frequency with which they occur in the raw material mix.
Trace Elements
Blaine et al. (1965) regarded the elements occurring at less than 0.02% each as the
“trace” elements. According to Sprung (1988), elements presents at levels less than 100
ppm are classified as trace elements. Because of their extremely small concentration
levels, it seems unlikely that the presence of trace elements will have any significant
effects on cement manufacturing. However, their effects on clinker can significantly
change if concentrations are increased beyond certain levels.
For the sake of convenience, the terminology “minor elements” has been used
throughout the text to cover both minor and trace elements, as defined by Blaine et al.
(1965) and Rompps Chemie-Lexikon (1987) respectively, unless mentioned otherwise.
Emphasis in this report is given to the minor and trace elements because of their likely
presence not only in the wastes but also in the conventional raw materials, and their
potential influence on cement manufacturing and use. It may be pointed out that trace
elements in a raw feed at one cement plant could significantly differ from another. As an
extreme, lead content in one plant may be 100-500 ppm compared to only 1 ppm in
another plant (Chadbourne, 1990).
Minor elements in cement primarily come from the raw materials and fuel used in cement
making. Examples of these are limestone, clay/shale, and coal. They also come from the
widely used auxiliary materials such as blast furnace slag, fly ash, silica sand, iron oxide,
bauxite, and spent catalysts. A secondary but important source of minor elements comes
from the wide range of industrial by-products which are partially or totally being
substituted for the primarily fuel. These include petroleum coke, used tires, impregnated
sawdust, waste oils, lubricants, sewage sludge, metal cutting fluids, and waste solvents,
as listed in Table 2.
Minor compounds found in several raw feeds for cements manufacturing as quoted by
Bucchi (1980) are shown in Table 3. Similar data on major components of raw materials
are shown in Table 4 and 5. They are limestone and shale/clay; widely used auxiliary raw
minerals, i.e. blast furnace slag (used up to 30% by weight of raw material), and coal fly
ash (used up to 15% by weight). Minor elements found in conventional kiln fuel (coal),
along with two secondary fuels (used oil and petroleum coke) are shown in Table 6.
Average values of minor components found in typical clinkers (Moir and Glasser, 1992)
are given in table 7.
Although blast furnace slag can be used up to 30% by weight, the level of use may be
reduced due to its magnesium oxide (MgO) content, particularly if the MgO level is
already high in the other raw materials. Bauxite is reported to contain 2-8% titanium
oxide (TiO2) and 0.04-0.4% chromium oxide (Cr 2O3). Iron ores frequently contain
chromium, arsenic, cadmium, and thallium, and may have adverse environmental
consequences because of their toxicity characteristics. A list of metals having regulatory
and environmental concerns has been specified by waste characterization regulations
under the Resources Conservation and Recovery Act (RCRA) and the Boiler and
Industrial Furnace (BIF) regulations that control treatment of hazardous waste in cement
kilns. These are shown in table 8 (Klemm, 1993). Both RCRA and BIF regulations apply
to wastes and require the use of the TCLP (Toxicity Characteristic Leaching Procedure)
tests.
The levels of sulfur and chlorine in bituminous coal, the main fuel for cement kilns, varies
from 0.5 to 4% and 0.007 to 0.39%, respectively. In some Illinois coals, sulfur is present
up to 6% by weight. Petroleum coke, used as an auxiliary fuel, contains up to 5% sulfur
and 0.6% vanadium oxide, and can contribute certain levels of S and V to clinker when
supplementing for coal. Tires have a zinc content of 1.2-2.6%. However, if tires replace
10% of the primary fuel, the resulting zinc oxide (ZnO) contents in clinker are increased
by only 0.03% (Sprung, 1985).
Additional sources of minor components could be the refractories, chains, and the
grinding media such as liners and grinding balls. A damaged chrome refractory lining can
enter into the incoming raw mix and incorporate a detectable amount of chromium into
the clinker. Partly for this reason, and mostly because of problems with their safe
disposal, the use of chrome bricks is being phased out in most parts of the world.
* Raw material includes natural materials such as limestone, clay, shale, sand, etc.
The role of minor and trace elements in the formation of clinker and their effect on
cement properties are discussed in this report as per their occurrences in raw mixes.
The elements chosen for discussion are categorized according to the periodic table, as
highlighted in Figure 2. They are discussed in their increasing order of atomic number.
The presence of any information gaps are identified and referred for further investigation.
ELEMENTS IN GROUP I
(Hydrogen, Lithium, Sodium, Potassium, Rubidium, Cesium)
Hydrogen
The role of Hydrogen (H) in cement manufacturing has not been documented in detail,
because hydrogen per se does not exist for long in the kiln as hydrogen is highly
combustible. It is present in the Kiln as water vapor, which results from the evaporation
of physically bound moisture from the raw material, and from the evaporation of water
sprayed on the raw feed to control dust during processing. Water vapor can also be
present in the kiln gases from combustion of fuel such as CH4 + 2O2 CO2 + 2 H2O
(Hawkins, 1994; Miller, 1994). A portion of water may also come from the dehydration of
raw materials such as clays, where it can be present in significant quantities depending
upon their mineralogical nature. In a wet plant, water comes from slurry.
The water vapor present in the kiln might have an indirect effect on the volatility of
alkalies which can increase with vapor pressure at higher temperatures, for example:
Apart from that, hydrogen may not have significant effect on the anhydrous nature of
clinker.
As a point of information, it may be mentioned that early cement kilns sometimes used
“producer gas” as fuel. The gas was generated (as a mixture of CO and H2) by the action
of steam on hot coal or charcoal as follows:
H2O + C CO + H2
Lithium
Lithium (Li) is found in some waste materials such as used lubricants, but occurs only in
traces in the kiln raw feeds and common fuels.
Lithium might behave somewhat differently from sodium or potassium in that it would
tend to form a relatively nonvolatile oxide (LiO 2) at elevated kiln temperatures. Gouda
(1980) reported that LiO2 is most reactive in lowering the temperature of the initial liquid
phase; the effectiveness has been shown as LiO2>Na2O>K2O. The presence of LiO2 also
disturbs the course of the burning process during which lime dissolves in the liquid
phase and results in higher reactivity. As a negative effect, LiO 2 also inhibits the
conversion of C2S to C3S. The effects are more pronounced with LiO 2 compared to Na2O
and K2O.
Rangarao (1977) pointed out that up to 1% of LiO 2 in the raw mix imparts a mineralizing
effect, but lime fixation is impaired afterwards. At a lower LiO2 addition (0.1-0.3% by
weight), limestone dissociation activation energy is reduced, and mineral formation
becomes more intensive.
If present in adequate amounts as an admixture, Li can have beneficial effects on
cement properties, since it is known to greatly reduce the alkali-silica reaction (ASR) in
concrete. Recent studies by Stark et al. (1993) have demonstrated that Li salts like LiOH
and Li carbonates (if added in appropriate amounts) reduce the ASR significantly. It is
conceivable that Li in clinkers could also have the same affect on ASR.
Since both sodium (Na) and potassium (K) occur together in raw feed, and by virtue of
similarities in their behavior in cement manufacture, it is appropriate to discuss them
together.
Sodium and potassium are mainly derived from the raw materials; their main carrier is
clayey rock. Sedimentary rocks, including the carbonate ores, sometimes contain soluble
alkali salts. Lea (1971) has quoted the occurrence of Na and K (by weight %) in different
components of raw materials used for cement manufacturing as shown in Table 9.
% by wt. Na2O K2 O
Typical raw mix 0.13 0.52
Limestone 0.26 0.11
Chalk 0.09 0.04
Marl 0.12 0.66
Clay 0.74 2.61
Shale 0.82 4.56
Alkalies frequently occur in auxiliary raw materials such as blast furnaces slag and fly
ash as shown in Table 5. Na2O and K2O in European fuels range between 0.05-0.6%
and 0.5-2% respectively (Bucchi, 1980); for typical raw feed, their average concentration
levels are 0.17% and 0.57%, respectively, as shown in Table 3.
Jawed and Skalny (1977,1978) and Skalny and Klemm (1981) have reviewed in detail
the effects of alkalies in cement manufacture and use. Na 2O and K2O are volatile in
nature, giving rise to a cycle inside the kiln. The extent of alkali volatilization varies with
raw material composition; for instance, volatilization of alkalies in clays is higher than
those found in feldspar. About half of the total alkalies by weight in the feed are
volatilized between 800-1000°C as the mix nears the burning zone, but condenses at
cooler parts of the system, such as in suspension preheater riser ducts or in chain
systems in dry kilns. The formation of rings and coatings on kiln lining resulting from this
heating-cooling cycle are generally attributed to alkali condensation and reaction with
refractories or incoming material. To avoid excessive buildups in the kiln or preheater
vessels, a percentage of gases may be bled through a by-pass, so that the alkali
sulfates and chlorides may be continuously removed and end up in the cement kiln dust
(CKD). Thus, CKD’s collected from by-pass dust collector are typically high in alkali
contents. Usually potassium compounds are more volatile than the sodium compounds.
According to Bucchi (1981), the intensity of the alkali cycle depends upon the nature of
their presence in raw material, on operating practices, and on type of kiln. The retention
of alkalies in clinker is generally higher for high efficiency kiln systems (Lea,1971). With
gas and oil as fuels, the alkalies tend to volatilize more as compared to coal used as a
fuel. This may be due to the high intensity of the flame with oil and gas compared to
coal.
In the presence of chlorides and sulfate, the volatilization behavior of both Na and K is
modified greatly, as shown by the vapor pressure-kiln temperature relationship in Figure
3. The vapor pressures of alkali carbonates resemble those of sulfates and they exhibit
similar effects. In the presence of sulfur, alkalies preferentially form sulfates. If their
amount is more than the required stoichiometric balance, the excess will be dissolved in
the silicates, aluminates and ferrites. The common alkali sulfate phases formed are
K2SO4, also known as arcanite, sodium potassium sulfate, also known as aphthitalite of a
general solid solution composition (K, Na) 2SO4 (commonly written as K3NŠ4 in cement
chemist’s notation), and Na2SO4, also known as thenardite (Taylor, 1990).
30
800
600
20
PCA Research
400and Development Bulletin RD109T 16
200 10
Role of Minor Elements in Cement Manufacture and Use
KCl
Vapor Pressure,
Vapor Pressure, mmHg mmHg (Sulfates)
(Chlorides)
NaCl
Na2SO4
K2SO4
According to Skalny et al. (1981), the ratio of Na to K in cement raw materials in the
North America and Europe varies. There is usually a substantial excess of K2O over
Na2O. Therefore, in the presence of sufficient amount of SO 3 a range of double alkali
sulfates, as described later in this report, is formed depending upon the K 2O to Na2O
ratio. If K2O is in excess of that required to produce aphthitalite, it forms arcanite.
Burning conditions also significantly influence the formation of sulfate so that oxidatizing
conditions produce calcium-potassium sulfate and a reducing condition produces
sodium-potassium sulfate. Potassium is twice a likely to produce soluble sulfates as
sodium. According to Pollitt and Brown (1968), the calcium potassium salt, calcium
langbeinite 2CaSO4.K2SO4 (also written as 2CŠ.KŠ), is also sometimes found.
Introducing SO3 jointly with K2O and Na2O into clinker melt leads to phase separation.
Since alkalies reduce the melt temperature, and the rate of C 3S formation is proportional
to the amount of liquid phase, a positive effect on C 3S formation could be expected.
However, Johansen (1977) reported that C3S, with or without the presence of alkali, has
the same amount of free lime after firing at 1400-1500°C. Alkali sulfate melt and clinker
liquid are immiscible phases. Alkalies inhibit the formation of C 3S from C2S and lime by
stabilizing lower energy C2S in the absence of sulfates.
After allocating for the sulfates, the remaining alkalies are distributed between silicates,
aluminates, and aluminoferrite. Lea (1971) has reported ranges of alkalies in the major
clinker phases shown in Table 10. The values are in general agreement with those
quoted by Taylor (1990) on the distribution of alkalies in different clinker phases.
Gartner (1980) and Gies et al. (1986) reported that in the absence of SO 3, Na2O is
preferentially incorporated in C3A by replacing CaO and forms “alkali-aluminate” of an
approximate composition NaO.8CaO.3Al2O3 (commonly written as NC 8A3), thereby
reducing its reactivity. This results in clinker rich in free lime and aluminate, and can
The mineralizing effectiveness of alkalies (in terms of decreasing melt viscosity, and free
lime contents in clinker) also appears to be a function of their cation size,
electronegativity, or the ionic potential. Such relationships for K, Na, Li and other relevant
cations are given in Table 11 (Grachian et al.), and in Figure 4 (Teoreanu and Tran van
Huynh, 1970) respectively. It might be mentioned that although alkalies, Na 2O in
particular, may act as fluxes, they are technically less desirable compounds than many
of the other available minor compounds (Bucchi, 1980).
If present in excess, alkalies often lead to higher pH and better early strength, but lower
later strengths. They are not desirable because of their deleterious alkali-silica reaction
(ASR) with reactive aggregates that leads to expansive reactions and can cause serious
cracking in concrete. ASR can be prevented with proper use of pozzolans.
Butt et al. (1971) reported that the deleterious effects of alkalies on the mechanical
properties of cement may be reduced by gypsum addition to the raw feed. They
considered this because of the possible elimination of solid solutions of alkalies with
clinker minerals. One possible speculation derived from the microscopic studies by Prout
(1985), is that gypsum would either increase the volatilization, or eliminate
NaO.8CaO.3Al2O3 or K2O.23CaO.12SiO2 formation. According to Lokot et al. (1969), the
addition of gypsum to raw feed produces cement of high 28-day strength, enhancing kiln
output and fuel savings.
7
[CaO]o = Free CaO in the Absence of Mg+2
Mineralizer (%)
4 Lia+1
Na+1
K+1
3
The remaining Group I elements, rubidium (Rb) and Cesium (Cs), are found only as
traces in cement raw mix or in the fuel. Rubidium generally occurs in cement at 0.01% or
less (Blaine, 1965).
Both Rb and Cs are expected to behave similarly to Na and K, in that they would both
form stables sulfates and volatile chlorides in the kiln (Gartner, 1980). On the other hand,
their concentrations may be too low to effectively influence the clinker formation or
cement properties.
ELEMENTS IN GROUP II
(Beryllium, Magnesium, Calcium, Strontium, Barium)
Beryllium
Beryllium (Be) would be present only in trace amounts in the raw feed and fuel (see
Tables 4 and 6). It is found only occasionally in the fine fractions of fly ash, an auxiliary
material frequently used as substitute raw material. Beryllium is found at a 55 ppm level
in =4m fraction compared to 12 ppm in > 45 m fraction of fly ash (Davison et. al.,
1974).
It may be suggested that because of its low volatile oxides, beryllium would stay in the
clinker. Nonetheless, beryllium has not been measured in significant amounts in clinkers
to have any measurable effects on clinker formation or cement use. Beryllium in today’s
cements has occurred up to 3 ppm (PCA, 1992).
Magnesium
Magnesium salt solutions (sulfate and chloride) are aggressive towards concrete and
react with the calcium hydroxide phase to form basic salts. The reactions are expansive
and may lead to deterioration of concrete.
Calcium
The role of calcium (Ca) in cement manufacturing has already been dealt with in the
section of major elements.
Strontium
A major portion of strontium (Sr) found in clinker as SrO comes from limestone and
aragonite.
Strontium as SrO, frequently occurs in clinkers. The mean value quoted by Moir and
Glasser (1992) in table 7 is 0.09%. Brisi et al. (1965) and Gilioli et al. (1972, 1973),
demonstrated that small amounts of SrO favor alite formation, but at 4-5% addition, Sr
preferentially distributes in belite rather than alite. Sr in belite inhibits alite formation.
Phase equilibrium studies indicate that Sr in raw feed also favors free lime formation,
with SrO preferably going into solid solution and displacing CaO from other compounds.
The tendency of free CaO release during clinkering makes SrCO 3 more labile than
SrSO4, and the clinkers having a high lime saturation factor (LSF) may be more
vulnerable to free lime expansion during hydration.
Butt et al. (1968) reported that the hydraulicity and strengths developed by Sr-doped
alites are significantly lower than the normal alites. This may be attributed to the smaller
sizes of the lattice voids in strontium-incorporated alite.
Kantro (1975) reported a slight set acceleration effect with SrCl 2.6H2O used as an
admixture in C3S paste.
Barium
Barium (Ba) occurs in varying amounts in limestones, mostly as barite (BaSO 4). It can
also occur in clayey sediments in appreciable amounts. The average amount of barium
in cement is 280 mg/kg. The average for CKD is 172 mg/kg (PCA, 1992).
Barium can be present in used oils. Excessive amounts in raw mix can increase the free
lime content of clinker due to CaO displacement and can cause expansion in concrete
under certain circumstances. It can also lead to paste shrinkage.
Boron
Boron (B) is generally found in traces (3 ppm) in most cement raw materials, particularly
those containing iron ore.
From early studies by Mircea (1965) it appears that B 2O3 reacts with C3S to form
C2S,C5BS (B is B2O3 in C5BS), and free lime. Upon further addition of B 2O3,C3S
completely disappears. Timashev (1980) established a relationship between the
electronegativity of boron and the melt viscosity, and noted a similarity between borates,
phosphates, and sulfates. Boron inhibited the formation of C3S and affected the stability
of the other major clinker phases. In the presence of boron C3S is decomposed to a
stabilized C2S as follows:
It was also pointed out that although B2O3 may not be a useful addition for regular alite
clinker required for early strength development, it might be useful as a mineralizer for
clinkers rich in belite. Gartner (1980) has reported on the effectiveness of B2O3 to
stabilize -C2S and to improve its hydraulicity. According to Miller (1976), boron can also
stabilize -C2S in alumina and iron-poor systems.
However, Miller (1976) has cautioned that the indiscriminate addition of boron can
produce unpredictable hydration results. Gartner (1980) explained that this behavior of
boron is probably sensitive to the presence of other trace elements. Bozhenov et al.
(1962) reported that even small additions of B 2O3 (~0.04%), as an admixture, to cements
can have adverse effects on setting properties. These observations indicate that B2O3 is
a strong retarder of cement hydration.
Aluminum
Role of aluminum (Al) in cement manufacturing has been dealt with in the section of
major elements.
Gallium (Ga), Indium (In), and Thallium (Tl) are found only in traces in raw material; their
typical concentrations in coal are 5-10 ppm, 0.07 ppm, and 1.1 ppm respectively.
Thallium and gallium are also found sometimes in the coal fly ashes. Thallium may also
be found in some pyritic minerals used as an iron source for raw feed. The average
concentration of thallium in cement is 1.08 mg/kg, ranging from nondetectable to 2.68
mg/kg. The average concentration of thallium in CKD is 43.24 mg/kg (PCA, 1992).
Although thallium occurs in traces in the raw feed, it is the most volatile element (4) after
mercury in the kiln (melting point=303°C), and is most likely to concentrate in the kiln
dust. The volatility of Tl relative to other elements in the kilns is shown in Figure 5
(Sprung et al., 1984). Sprung et al. determined the volatility on the basis of the difference
between the external and internal balances of individuals elements during the clinker
burning in a cyclone preheater kiln. Indium is also volatile and largely ends up in the kiln
dust. Since thallium may concentrate in the fly ash from the coal firing power plants, in
the cement kiln operation it tends to build up in extremely large internal cycles if no dust
is discarded. 957
(4)
Nonvolatile elements are often called refractory elements.
150
100
* Relative
Volatility (%)
50
0
Ca Zn V Be Ni K Cr As Pb S Cd Cl Tl
* Relative volatility as a percentage of ratio between the total external and internal balance for a given element.
ELEMENTS IN GROUP IV
(Carbon, Silicon, Germanium, Tin, Lead)
Carbon
Carbon (C) is a major component of fuel. It is also present as carbonate in the limestone.
A significant amount of carbon can also present in fly ash as unburnt coal.
Carbon as CO2 is extensively present in cement kiln systems, but is not present in any
significant levels in clinker. Because of the limestone and fuel that are used in the kiln,
the gasses emitted from the kiln systems are constituted mainly of CO 2, H2O, and N2.
Limestone (CaCO3) decomposes to CaO and CO2 at about 900°C. Roughly for every ton
of clinker, one ton of CO2 is generated in the kiln, which essentially is released through
stack emissions.
Silicon
Role of silicon (Si) in cement manufacturing has already been discussed in the section of
major elements.
Germanium
Germanium oxide (GeO2), is not volatile (Gartner, 1980), and is likely to concentrate in
clinker. When present in large amounts, GeO2 can form C3G*, tricalcium germanate with
CaCO3 at 1500°C and is stable between 1335-1880°C. At temperatures below 1335°C,
C3G decomposes to C2G and free lime (Hahn et al., 1970; Boikova et al., 1974). These
forms of calcium germanates are similar to C 3S and C2S respectively. C3G is hydraulic
and produces calcium germanate hydrate (C-G-H) and calcium hydroxide (CH) with
water, whereas C2G is assumed to be non-hydraulic. According to Gartner (1980), it is
unlikely that the trace amounts of Ge would seriously affect the formation of clinker and
the properties of the resulting cement.
Tin
Tin (Sn) is a trace element in both the raw feed and fuel. Tin is reasonably nonvolatile
(boiling point=2265°C). Tin oxide (SnO) or natural cassiterite melts at 1630°C and
sublimes between 1800 and 1900°C. It is very likely that tin will stay in the clinker. The
presence of trace amounts of tin in clinker should not affect cement properties, although
not much is known about the effect of tin in clinker manufacture.
Lead
Lead (Pb) can be present in trace amounts in raw material mainly in clay and shale. It
would be present at appreciable levels in coals, used oils, lubricating oils, and scrap
tires. In fly ash, lead tends to concentrate in the fine fractions (Coles, 1979). Lead levels
in coal, used oil, and petroleum coke are shown in Table 6. Another source of lead could
also be the lead shot from shot gun shells used to shoot out rings.
* G= GeO2
The effect of lead in cement manufacturing and properties has been studied in some
detail. Lead compounds are fairly volatile. They tend to vaporize in the kiln, and exit the
kiln as fines and are collected in the kiln dust.
There is also evidence that despite the partitioning of lead into the CKD, some lead can
still be retained in the clinker (Davison et al., 1974; and Berry et al., 1975). However, Pb
has been shown to have no adverse effect on cement properties if present below 70
ppm. The effect of lead levels higher than that in clinker is uncertain (Sprung et al.,
1978). According to a recent PCA study (PCA, 1992) the average lead levels in the
CKD’s and cements produced in North America are 434 ppm and 12 ppm respectively.
Some research on the effect of lead compound additions on hydrating cement properties
has recently been studied, where Bhatty and West (1992) have noted that additions
either as a soluble compound (PbNO3:7,300 ppm level) or insoluble oxide (PbO:38,000
ppm level) substantially retards the hydration of pastes, but enhances the workability.
The retardation effects are more pronounced with oxides. The initial setting time is
increased with a consequent loss in early strength, but the 28 and 90-day strengths are
comparable to or higher than those of the control.
ELEMENTS IN GROUP V
(Nitrogen, Phosphorus, Arsenic, Antimony, Bismuth)
Nitrogen
Nitrogen (N) can be present up to 0.01% by weight in the raw materials, but in coal and
other fuels nitrogen can be as high as 1-2%, often as hetrocyclic nitrogen compounds.
Clinker made under reducing conditions tend to have up to 0.05% N as nitrides. Under
normal oxidizing conditions, nitrogen in clinker is present only at a few ppm.
The quantity of thermal NO formed is closely related to the burning zone temperature
(BZT). According to Lowes et al. (1989), a reduction in BZT from 1500°C to 1300°C can
reduce the NOx levels by 200-400 ppm. Nitrogen in coal or other fuels, present at about
the 1-2% level, is considered significant in producing NO emissions from cement plants.
However, it is not known to the degree in which the nitrogen in the kiln raw feed also
contributes to NOx emissions (Gartner, 1980). In precalciners, fuel nitrogen may play a
role, but in the burning zone the temperature is so high that thermal NO is virtually in
equilibrium.
Phosphorus
Phosphorus (P) as phosphates is present in limestone and shale (Moir et al., 1992); they
are also present in sandstones, sands, and in detrital clays (Bucchi, 1980). Phosphorus
also occurs in the blast furnace slags, electric furnace slags, convector slags, and fly ash
which are often used as substitute raw feed for cement manufacturing. Phosphate is
found in sewage sludge which is a potential partial kiln fuel.
Cement clinkers contain typically around 0.2% P 2O5 (Lea, 1971). A high P2O5
concentration decomposes C3S to C2S and excess lime. If P2O5 is present in excess of
2.5% by weight, the formation of free lime occurs (Nurse, 1952). However, by correct
proportioning and proper burning, sound clinker can be produced, but cement hardening
becomes slower. Matkovich et al. (1986) reported higher hydraulic actively for ´C2S
stabilized by P2O5 than for the -C2S.
In a CaO-C2S-C3P* system at 1500°C, raw mix with more than a few percent P 2O5 does
not yield C3S. However, in the presence of fluorine, the tolerance to P 2O5 is somewhat
improved. It is very likely that the thermodynamics of the system favor the fluoride-
aluminum-C3S solid solution rather than P-C2S solid solution (Gurevich et al., 1977) and
apparently form a fluoroapatite phase (100CaO.3P2O5.CaF2) which is dissolved in C3S.
Gartner (1980) suggested that chlorides may also help stabilize P 2O5 in C3S by forming a
stable chloroapatite (10CaO.3P2O5.CaCl2) which also forms a stable solid solution with
fluoroapatite.
Coleman (1992) reported that an appropriate level of P 2O5 in clinker reduces the
negative effects of alkali on strength properties of cements. He reported that in cement
clinkers with “normal” Na2O contents of 0.8%, the maximum 28-day strength was
achieved at 1.0% P2O5 level.
Arsenic
Arsenic (As) bearing mineral arsenolite or claudite As 2O3 (or As4O6), occurs only in small
amounts in coal and used oils, and are unlikely to influence cement manufacturing in any
way. Smith et al. (1979) have indicated that in coal-fired power plants, As tends to
concentrate in the fly ash, but its concentration level, as detected by the XRF method, is
extremely low. It tends to concentrate in the fine fractions of fly ash where the levels can
go up to 70 ppm. Weisweiler et al. (1989) has reported up to 5 ppm of As in raw material
and only 0.6 ppm in petroleum coke. Arsenic levels found in various materials are shown
in Tables 4-6. The average concentration of As in cement and CKD is 19 mg/kg and 18
mg/kg respectively (PCA, 1992).
*P=P2O5
As a metalloid occurring in different oxidation states, arsenic can have complex effects
on the hydration properties of cement (Conners, 1990). Tashiro et al. (1977) reported
that As2O3 only slightly retards the paste hydration when added up to 5%. It was found
that the As leaching rate from hardened cement mortars using either ordinary water or
sea water, although measurable, was very low.
Antimony
Antimony (Sb) occurs as traces in cement raw materials. It has been reported to occur at
0.08 ppm in the raw feed and 0.0429 ppm in petroleum coke (Weisweiler et al., 1989),
Sprunge (1985) has quoted 1.19 ppm Sb in coal. According to measurements in BIF
certification of compliance (C.O.C.) and other authors, the Sb levels in raw materials are
higher.
Bismuth
Bismuth (Bi) occurs as a trace element in the raw feed and fuel. The stable oxide Bi 2O3
is not volatile at clinkering temperature (boiling point=1860°C). Little is available on the
influence of Bi in cement manufacturing and cement hydration, but owing to trace
concentration, it is conceivable that the effects will be practically insignificant.
ELEMENTS IN GROUP VI
(Oxygen, Sulfur, Selenium, Tellurium)
Oxygen
The role of oxygen (O) per se on the manufacture and use of cement has not been
studied. Nonetheless, a considerable portion of raw material and clinker phases
incorporate oxygen in one form or the other. Raw material is primarily composed of
CaCO3 (~75%), SiO2 (~20%), and Al2O3 (~2%). CaCO3 in the raw mix is derived from
limestone; SiO2 and Al2O3 from clays, shales, sandstones, bauxite, and Fe2O3 from iron
oxides and iron ores.
The importance of oxygen levels is also related to the effect on the environment of the
kiln and the kind of reactions that are favored. Thus, the presence of oxidizing or
reducing atmospheres greatly influences the reaction into which the various elements
will enter. Clinker made under oxidizing conditions tends to incorporate trace metals of
higher oxidation states than clinker prepared under reducing conditions. Examples of
chromium and sulfur can be cited here. Cr+6 would tend to form under oxidizing
conditions, instead of Cr+3, which results under reducing conditions. Cr has also been
reported to occur as Cr+4, Cr+4.6, and Cr+5, (Johansen, 1972), but eventually they
disproportionate to more stable Cr+3 or Cr+6 when mixed with water.
Alkali sulfates formed in the kiln are preferably decomposed under reducing conditions.
Kilns having strongly oxidizing conditions and low burning zone temperature tend to
retain more sulfur in clinker than those produced under reducing conditions and for high
burning zone temperature. Thus, the oxidation or reducing condition in the kiln can lead
to significant phase modifications in clinker.
Clinker produced under reducing conditions are brownish as compared to darker gray
clinkers made under oxidation conditions, most probably because of the oxidation state
of iron. Burning conditions may also have an effect on the crystallinity of major phases.
The effects can be pronounced if trace metals are also present.
Sulfur
Sulfur (S) is frequently present in coals and some fuel oils; sulfates and sulfides are also
often present in the limestones. Clayey sediments, marls, also contain both sulfides and
sulfates. Locher et al. (1972) have reported occasional use of gypsum and anhydrite as
mineralizers and modifiers of the alkali cycle in the kiln.
Sulfides and sulfur from raw material and fuel are oxidized and are incorporated into the
solid phases as sulfates in the clinker, though some sulfur as SO2 will almost always
escape with the existing gases.
Sulfur forms volatile compounds and its behavior in a kiln is a complex one. Depending
upon the burning conditions in the kiln, both oxidized and reduce species may occur in
solid, molten and vapor phases, as explained by Choi et al. (1988) in Figure 6. Under
oxidizing conditions at high temperature, the formation of SO 2 is most likely. In the
presence of lime, SO2 is partly removed to form CaSO4 by the following mechanism:
Tempera
I ture Possible carry-through SO2 prominent Molten sulfates
n of complex calcium in vapor pressure
c sulfides in clinker
r
e
a S-2 present as organic Sulfites, SO3-2 in solids Sulfate solids
s and inorganic forms in which become SO3 vapors
i fuel, etc. increasingly unstable
with rising temperature
n
g
Reduced S Species Intermediate S Species Oxidized S Species
Figure 6. Formation of different sulfur species in cement clinkering (Choi and Glasser, 1988).
In the presence of alkali, alkali sulfates are formed which are later condensed at the
lower temperature regions. These condensates, from liquids and solids, contribute to
build up problems in various kiln systems. Intermediate compound such as
“sulfospurrite” 2C2S.CŠ, and the ternary compound “sulfoaluminate” C4A3Š also
condense at lower temperatures.
Another well known problem of sulfur being volatile is its cycle of vaporization and
condensation with alkalies. They are volatilized at high temperatures and subsequently
condense on the relatively cooler incoming raw feed resulting in high sulfur and alkali
levels in the middle zone of kiln, especially with preheaters. The use of an alkali by-pass
is often effective to break this cycle and lead to the reduction of sulfur and alkalies in the
incoming kiln feed. However, alkali sulfate levels are significantly increased in the by-
pass dust, which is captured by the dust-collector and generally discarded.
Sulfates preferably combine with alkalies to give alkali sulfates in clinker as (K, Na) 2 SO4,
known as aphthitalite, or K2SO4 known as arcanite. If sulfate is present in excess, the
balance between alkali is achieved by forming calcium langbeinite, Ca2K2(SO4)3, which is
stable up to 1011°C in a CaSO 4-K2SO4 system. However, this phase is known to
evaporate incongruently at high temperatures, and vaporizes K and S (Arceo et al.,
1990).
Major alkali salts formed with sulfates and their approximate melting temperatures
according to Gartner et al. (1987) and Skalny and Klemm (1981) are shown in Table 12.
Strunge et al. (1985) reported that increasing sulfate contents distinctly decreases alite,
increases belite; the aluminates and ferrite contents are unchanged in clinkers
irrespective of their silica modulus (SM) values. On the other hand with increasing SM,
irrespective of the sulfate, the alite contents are higher, belite is unchanged, and
aluminates and ferrite are somewhat lower. Relationships between clinker phases and
sulfate content in the clinker are shown in Figure 7. With increasing sulfate contents, the
alite crystals in clinker grow larger, and the tendency of belite inclusion in alite is
progressively reduced. The crystal size of aluminate and ferrite phases are also
significantly reduced.
Table 12. Major Alkali Sulfates Formed During Clinkering and their Approximate Melting
Temperatures (Adopted from Skalny and Klemm, 1981; and Gartner et al., 1987)
80
SM = 1.6 SM = 2.4 SM = 3.2
Alite Alite
60
Alite
Clinker
Phases %
mass
40
Belite
Belite Belite
20 Aluminate Aluminate
Aluminate
0 1 2 3 0 1 2 3 0 1 2 3
SO3 Content, % mass
Another related concern is the level of SO2 in the kiln exhaust area. Very frequently, 15-
40% of pyritic (sulfide) sulfur in raw material is converted to SO 2 in the emissions
(Neilson, 1991).
It should be pointed out that in the preheater system much of the SO 2 in the kiln is taken
up by the incoming raw material. This reaction is also observed in plants which use kiln
exhaust to provide heat to the raw milling system. Significant amounts of SO 2 may still
escape if its original concentration is high, or if reducing conditions are generated locally.
Selenium
Selenium (Se) could be associated with sulfur in coal, but only in traces. It is also
present in fly ash where it tends to concentrate in the fine fractions (Coles, 1979).
Selenium is usually not detectable in cement but is detected in CKD in small amounts
(PCA, 1992).
Selenium is volatile (boiling poin=684°C) and expected to end up in kiln dust or in the
emissions. Selenium could form less stable selenates (SeO 4), which are unlikely to stay
in clinker (Gartner, 1980). Since their concentration is extremely low in the kiln feed, it is
very unlikely that they will have any significant effect on the manufacture or properties of
cement.
Tellurium
Like selenium, traces of tellurium (Te) are generally associated with sulfur in coal.
At optimum kiln temperature tellurium could be somewhat volatile depending upon the
form in which it is present (amorphous form boiling point=990°C; rhombohedral form
boiling point=1390°C). Gartner (1980) suggests that tellurium might form unstable
tellurates in clinker and end up in the kiln dust or the emissions.
The halogens fluorine, chlorine, bromine, and iodine, are frequently found in kiln raw
feed and primary as well as alternative fuels, and therefore play an important role in
cement manufacturing. Some halides such as fluorides are also frequently used as
mineralizers in clinker production and in low-temperature manufacturing of belite-rich
cements. Mishulovich (1994) addresses halides as catalysts for calcination.
Concentration of halogens found in raw materials and fuels is given in Tables 3,4 and 6.
Fluorine
Fluorine (F) is commonly present in limestone, clay/shale, and coal (Sprung et al.,
1968,1985) as a minor element and plays an important role in cement making. In raw
feed, fluorine could be up to 0.06% by weight (see Table 3), whereas in limestone and
clay /shale it can go up to 940 and 990 ppm respectively (Table 4).
Calcium fluoride (CaF2) is also frequently added to raw meals as a mineralizer and flux
to lower the burning temperature and accelerate the formation of C 3S (Klemm et al.,
1976, 1979). Miller (1976) has, however, cautioned not to use fluoride beyond 0.25% to
avoid adverse effects on clinker behavior by selectively incorporating it into the
aluminates or silicates phases at certain burning temperatures. At lower temperatures,
fluoroaluminates (C11A7.CaF2) are formed, which are decomposed at high temperatures
to C3A and fluorides. These fluorides are then incorporated into silicates at higher
temperatures to form often stable fluorosilicates but their excessive amounts can cause
decomposition of alite.
Gartner (1980) also reported the formation of alkali fluorides as NaF and KF at higher
alkali presence; these fluorides beings somewhat volatile (boiling points 1700° and
1500°C respectively) are expected to end up in the fine kiln dust. However, Sprung et al.
(1968) reported that between 88% to 98% of fluorides are incorporated in the clinker and
only a small fraction en up in the kiln dust, probably as CaF 2. Fluoride emissions were
reported low (0.009-1.42 mg F/Nm 3) depending not necessarily on the magnitude of
fluoride balance but on the efficiency of the precipitators.
Akstinat et al. (1988) reported that fluorides have no adverse effects on the cement
production process, and the fluoride cycle does not cause any operational problems like
coating, because of their presence in small amounts. However, recent experiences have
shown that use of fluoride based compounds can occasionally cause plugging. Gartner
(1980) reported that the presence of fluorides beyond 0.5% can cause both operational
and quality control problems, which, under certain situations, can be controlled by P 2O5
addition. Goswami et al. (1991), Bolio-Arceo et al. (1990), and Gilioli et al. (1979), have
reported the formation of spurrite (2C2S.CaCO3), and fluor-ellestadite that cause kiln
deposits, but the resulting low burning temperatures control the alkali cycle and reduce
the alkali-sulfate deposits.
Palomo et al. (1985) suggested that 0.2% fluoride promotes low temperature formation
of aluminates such as fluorinated C12A7,C3A, and C2AS (gehlenite); however, the final
aluminate mineralogy was not significantly affected, as both ferrite and C 3A were present
at 1250°C and above. Perez Mendez et al. (1986) reported that with the addition of 0.5-
1.5% fluorides, as CaF2, clinkering reactions were completed in 0.5 hr at 1354°C; the
clinkers had much of C3S developed, with -C2S, C4AF and C3A also present therein.
Imlach (1974) observed that fluoroaluminate C11A7.CaF2 forms at fluorine levels of about
0.5% in clinkers fired below 1320°C or slowly cooled from 1340°C to 1265°C.
Fluoroaluminate imparts rapid setting to cement pastes compared to the normal
cements.
According to Aldous (1983), and Shame et al. (1987), the presence of F and Al beyond
the threshold level renders C3S a rhombohedral symmetry, which is associated with
improved hydraulic properties. Moir (1983) demonstrated that by optimizing the levels of
F, alumina, alkalies, and sulfates, the C3S in clinker could be maximized to enhance the
setting properties of cement, Figure 8 shows the relationship between the fluoride
addition and compressive strength of cement pastes at various curing ages. An optimum
fluoride addition for maximum strength at early ages (24 hours) was 0.2%, for later
strengths (7 days and 28 days) addition of 0.75% were acceptable.
600
Setting Time
(Minutes) 300
28-day
60
7-day
3-day
40
Compressive
Strength (N/mm2
)
20
24-hour
8-hour
16-hour
0
Figure 8. Effect of fluoride on strength and setting time of high alite cements (Moir, 1983)
Chlorine
The formation of stable yet volatile alkali chlorides NaCl (boiling temperature=1413°C)
and KCl (subliming temperature=1500°C) at clinkering temperature is well known. Both
the chlorides volatilize in the burning zone and condense in the cooler parts to form kiln
rings or preheater build-ups which impair plant performances. Bhatty (1985) also
concluded that agglomeration due to the presence of molten alkali chlorides was one of
the major reasons for the build-ups. Cl also enhances the formation of spurrite and
sulfospurrite (2C2S.CaSO4). In cases of plants without preheaters, the volatile chlorides
end up in the kiln dust. In preheater kilns, up to 99% chlorides are recaptured by the
incoming feed in the calcining zone (Ritzmann, 1971); the concentration of chloride at
that point could be extremely high (>1%) compared to that of raw feed (0.01%).
Relative volatility of Cl2 and other elements in the kiln system is already shown in Figure
5 (Sprung et al.).
The wet processing plant and grate preheaters may tolerate raw feed with higher
chlorides, but the limits primarily depend upon the efficiency of dust collecting and the
level of kiln dust recycling. With the advent of the alkali by-pass, the chlorine cycle can
be broken at the most intense point of kiln and the alkali chloride can be conveniently
directed to the dust collectors. Otherwise, as reported by Norbom (1973), a total chloride
intake of 0.015% (in both raw material and fuel) can result in build-ups in a preheater
without a by-pass.
Since most of the chlorides are volatile, the amount retained in clinker is extremely small
(<0.03%). Volatile chlorides react readily with alkalies, so that the alkali level in the
clinker is often reduced when chloride is present. The combined influence of alkali
chloride on cement properties is therefore regarded as insignificant. In some cases,
calcium chloride is added to the kiln for the express purpose of increasing alkali
volatilization and removal, and result in the production of a low alkali clinker. According to
Mishulovich (1994), the addition of calcium chloride and chlorine-containing organic
compounds at the clinkering stage accelerated both lime reaction and alkali volatilization.
In a preheater kiln, the addition of calcium chloride in the burning zone, resulted in 20%
increased production with corresponding fuel saving.
High chlorides in the raw feed have also been reported to form condensation plumes in
the emission stacks in long wet or dry kilns which are difficult to remove at times. Such
detached plumes are generally the result of NH4Cl formation. Excessive chlorides can
also have a deleterious effect on kiln basic brick lining.
Chlorides, particularly CaCl2, accelerate the hydration and hardening of cement paste
and increase the very early strength but, at the same time, chloride ions are also known
to promote corrosion of steel reinforcing bars in concrete.
Alinite cements: The development of less energy intensive “alinite cements” from the
CaCl2 incorporated raw material has generated great interest (Nudelman, 1980). The
formula ascribed to the alinite phase is close to 21CaO.6SiO2.Al2O3.CaCl2 with some
MgO inclusion (Locher, 1986). The burning temperature for alinite clinker is between
1000-1100°C. The raw mix is composed of 6-23% CaCl2 by weight. MgO is added to
stabilize the alinite phase at 60-80%, belite at 10-30%, calcium aluminochloride at 5-
10% and calciumaluminoferrite 2-10% (Bikbaou, 1980). Ftikos et al. (1991) reported that
the strength development of alinite cement was comparable to that of regular Portland
cement.
Bromine
With some exceptions, bromine (Br) plays a minor role in cement manufacturing.
Bromine occurs only as a minor element in raw materials, i.e. limestone (6 ppm), clay
(10-58 ppm), and coal (7-11 ppm (Akstinat et al., 1988 and Sprung et al. 1985). Bromine
has also been detected at measurable levels in some of the fly ashes generated at coal
operated power plants.
Bromine is volatile and expected to end up in stack emissions (Akstinat et al., 1988).
Under oxidation conditions, bromine gas (Br 2) would form and end up emissions.
Retention of bromine in clinker is negligible. Alkali bromides can also be found in cement
kiln dust. Between 5-10 ppm of bromine was reported in one CKD sample using fly ash
as a partial raw feed (Klemm, 1995). At higher levels of bromides, the formulation of
bromine-alinites analogous to chlorine-alinites, as mentioned above, has also been
reported by Kurdowski et al. (1987, 1989). Bromine-alinites are much more reactive than
the alites (Kurdowski et al. 1989).
Iodine
The presence of iodine (I) in limestone and clay is negligible. Up to 0.75 ppm in
limestone, 2.2 ppm in clay and shales (Mantus et al.), and between 0.8 to 11.2 ppm is
found in coal (Sprung 1985).
Because of the low levels of iodine in the feed, the effect on the burning process is
negligible. Iodine salts are volatile in nature and mostly end up in emissions. Conversion
of iodine gas (I2) from iodides is easier that bromides. Their concentration in clinker is
detected at very low levels. There is no literature report on iodine presence in the CKD.
The concentration of iodine in CKD is expected to be extremely low, may be in the ppb,
because of its presence in small amounts in the raw materials.
Helium (He), neon (Ne), argon (Ar), krypton (Kr), and xenon (Xe), being inert gases, are
not known to impart any noticeable effect on clinker manufacturing or cement hydration
properties.
TRANSITION ELEMENTS
The elements 21-30, 39-48, and 57-80 in the periodic table are known as the transition
elements. Not all of these elements have been studied in cement manufacturing, but the
ones that have been studied in some detail are dealt with in this section. Some of these
transition elements are introduced into the clinkering process through the use of spent
catalysts as an alumina source.
Yttrium
Isomorphism between yttrium (Y) and calcium frequently occurs in natural materials; for
instance fluoroapatite, Ca2Ca3(PO4)F, can contain up to 10.6% Y2O3 (Povarennykh,
1966). But presumably in cement raw materials, yttrium occurs only in traces.
Yttrium substitutes for Ca in both C3S and C2S (Boikova, 1986). It yields both triclinic and
monoclinic forms of C3S. In a C2S-Y4(SiO4)3 system, the region of homogeneity can exist
up to 35% Y4(SiO4)3 by weight (Toropov et al., 1962-2).
Titanium
Titanium (Ti) as oxide could be present in typical cement raw material at the 0.02-0.4%
level by weight (Bucchi, 1980). Gartner (1980) reported a higher concentration of 0.1-
1.0% TiO2 in many raw mixes. Concentrations of TiO 2 in some of the auxiliary raw
materials are even higher. Blast-furnace slag for instance, can contain 1.7% TiO 2, and
the bauxites may have between 2-8% TiO2 by weight.
Ti
V
Cr
Mn
Co
Ni
Cu
Zn
Weight %
Knöfel (1977) observed a sharp reduction in alite with equal gain in the belite phase
when TiO2 was increased in the raw mix; the variation in ferrite and aluminate was not
significant. Calcium titanate (CaTiO3) is apparently the major phase present in clinker. It
was also reported that about 1% TiO2 addition in the raw mix reduces the melt
temperature by 50-100°C, probably because of a favorable relationship between ionic
potentials and the melt viscosities as shown in Figure 10. This relationship was
developed by Timashev (1980). It shows that increasing the ionic potential of transition
elements in groups of elements with equivalent atomic radii decreases the clinker melt
viscosity.
20 W 600
Mo
V
400
15
W Cation-Oxygen
Ionic Potential to Bond Energy (KJ.
Radius Ratio 200 Avogadro´s
(1010.Va,m-1 ) V
Mo Number)
10
5 0
Viscosity, Pa sec
0
Figure 10. Relationship between viscosity and ionic potential to radius ratio, and
cation-oxygen bond of transition elements at 1450°C (Timashev, 1980)
Although TiO2 enhances the early hydraulicity of alite (Kondo, 1986), the clinkers have
shown slow initial setting. However, 1% TiO 2 clinker have roughly 20% higher 3-and 90-
day strengths (Knöfel, 1977, 1979).
Zirconium
Zirconium (Zr) is concentrated mostly in siliceous ores which can be used as a raw feed
component (Miller, 1976).
Blaine (1965) reported about 0.5% zirconium, probably in the fully oxidized form of ZrO 2,
in U.S. clinkers. Kakali et al. (1990) found no significant change in the burning and
cooling conditions for clinker prepared with 0.73-1.45% Zr2O3; the principal phases, alite,
belie, aluminate and ferrite, were satisfactorily crystallized. However, Zr2O3, changed the
size and shape of alite, while the type of belite crystal was modified. Zr2O3 also imparted
a noticeable color change in clinker (Kakali, 1988).
A significant retarding effect and a subsequent delay in strength for cements prepared
with ZrO containing raw mixes was also reported (Kakali et al., 1989). However, earlier
studies by Blaine et al. (1966) indicated that smaller ZrO additions increased the early
compressive strength of cement.
Vanadium
Vanadium (V) occurs at a measurable level in cement raw material (10-80 ppm in
limestone, 98-170 ppm in clay/shale, and 30-50 ppm in coal) (Sprung, 1985). It is also
present in fly ash where it tends to concentrate in the finer fractions (Coles, 1979). Fairly
high levels of vanadium are also reported in crude oils (Gartner, 1980).
In one study, Weisweiler et al. (1990) has reported nearly 800 ppm vanadium in
petroleum coke used in cement manufacturing. Ash from petroleum coke also contains
very high levels of V2O5 (up to 60%). Because the petroleum coke has a low overall ash
content, Moir et al. (1992) found no more than 0.08% V2O5 in clinker produced in modern
cement plants that use 50% petroleum coke as a substitute fuel.
Use of vanadium is known to decrease the melt viscosity primarily because of its higher
ionic potential as shown in figure 10. Vanadium is present as V 2O5 in cement clinker. It
concentrates in alite and forms larger crystals. However, according to Hornain (1971),
vanadium preferably concentrates in belite rather than alite, as shown in Figure 9. V 2O5
is unlikely to vaporize at normal kiln temperatures. On the other hand, vanadium present
in fuel may not have adequate contact with the reacting mass in the kiln and largely ends
up in the kiln dust as suggested by data from Weisweiler et al. (1990).
Odler et al. (1980-1) reported that 1% of V 2O5 can significantly reduce the free lime in
clinker when fired at 1200°C. Xinji et al. (1986) used V 2O5 for stabilizing -C2S in clinker
apparently by substituting VO4-3 for SiO4-4.
Niobium
Because of the low presence in the raw mix, niobium would have very little effect either
on the clinker formation or on the cement hydration properties. Kakali et al. (1990)
reported a very feeble effect of Nb+5 addition (up to 1.5% by weight) on the mineralogical
texture and the viscosity of clinker melts because of its low ionic charge to atomic radius
ratio.
Cement pastes prepared from these clinkers did not shown any noticeable change in
their setting or strength properties when compared to regular cement pastes (Kakali et
al., 1989).
Tantalum
Tantalum (Ta) is only a trace element in cement raw material. It reported to be present at
less than 9 ppm in raw material and 0.3 ppm in the oil coke used as fuel in cement
manufacturing (Weisweiler et al., 1990).
Since tantalum is present as trace in both the raw feed and fuel, it is unlikely to impart
any noticeable effect on the clinker formation and cement use. Weisweiler et al. (1990)
have reported 14.3 ppm and 3.3 ppm tantalum respectively in clinker and kiln dust
prepared from a raw material containing 8.9 ppm tantalum.
Chromium
Chromium (Cr) can be present in raw feed in measurable quantities. Sprung (1985) has
reported up to 16 ppm in limestone, nearly 100 ppm in clay and shales. Coals and used
oils may contain up to 80 ppm and 50 ppm Cr respectively. Some of the auxiliary raw
materials, such as bauxites, which are used up to 4% in cement manufacturing, may
contain between 0.04-0.4% Cr2O3. In addition to that, a proportion of Cr can also enter
cement from the grinding media during raw meal preparation and finished cement
grinding, and refractory linings.
The presence of Cr in raw materials is known to reduce the viscosity of clinker melt due
to its high ionic charge as is shown in figure 10. Miller (1976) has reported improved
clinker burnability at 1% Cr2O3 addition. Chromium can exist in a number of oxidation
states in clinker, the most stable being Cr +3 and Cr+6. Their formation is sensitive to the
oxygen level in kiln. High oxygen tends to form Cr+6 compounds as chromates, which are
readily soluble in water and markedly affect the hydration characteristics of the paste.
Reducing conditions favor the formation of Cr+3 compounds which are less soluble in mix
water.
Under oxidation conditions, Cr can also exist as Cr +4 and Cr+5 in C2S, which can then
disproportionate to the more stable Cr+3 and Cr+6 upon mixing with water (Feng Xiuji,
1988). Johansen (1972) has reported Cr+4, Cr+4.6, and Cr+5 in alite substituting for Si+4.
Hornain (1971) reported that Cr preferentially resides in belite followed by ferrite, alite
and aluminates, as shown in Table 14 (see also Figure 9). Although Cr +6 can be present
in both alite and belite, it is reported to be stabilizing the -C2S form (Hornain, 1971, and
Kondo, 1963). Subarao et al. (1987) developed an active belite-rich cement from raw
feed containing 4-5% Cr2O3 by weight. Imlach (1975) used 0.11-1.32% Cr2O3 in the raw
feed as a flux. The resulting cement exhibited improved 8- and 24-hour strengths, but
28-day strengths always decreased.
Phases Cr (%)
Belite 0.87
Ferrite 0.55
Alite 0.39
Aluminate 0.04
A significant portion of Cr can also enter the finished cement from chrome-rich grinding
media (Klemm, 1994). It is reported that the level of Cr +6 in ground cement is almost
doubled by the use of high-chromium alloy balls during grinding. A number of patents
report the use of inorganic reductants to control the Cr +6 leaching from cement. Most of
these patents are of European origin and use ferrous sulfate heptahydrate, ammonium-
ferrous sulfate, and manganese sulfate during intergrinding to convert Cr+6 to Cr+3.
Chromium is known to accelerate the hydration of paste and improve the early strength,
and has thus been used to develop high strength cements. Recent studies (Bhatty et al.,
1993) have shown that 0.75% addition of chromium as chromium chloride and nitrate,
accelerate paste hydration and result in high initial hydration peaks. The workability and
the initial setting times are reduced, but the early strengths (3 days) are significantly
improved over the control. The 28- and 90-day strengths are, however, close to those of
the control. The addition of insoluble chromium oxide (Cr 2O3), even up to 1.3%, did not
significantly affect the hydration or the strength behavior of pastes. The degree of Cr
stabilization in cement matrices as determined by leachability in both these cases was
almost 100%. Chromium may also contribute to high sulfate expansion, increased 24-
hour shrinkage, and reduce autoclave expansions.
Molybdenum
Tungsten
Tungsten (W) is trace metal in raw mix and is expected to appear in traces in the clinker.
Very little work has been reported on the effect of tungsten on clinker formation and use.
Kakali et al. (1990) noted that the addition of up to 1.5% WO 3 in the raw mix changed the
shape of alite crystals, making them bigger and more roundish; the belite formed was of
type III and, to some extent, contained secondary dendritic crystallization probably
because of excessive Si+4 replacement by W+6. Dissolution of W+6 in the melt decreased
the viscosity because of its large charge to radius ratio, as is also exhibited in Figure 10.
Ivashchenko (1991) reported that addition of W+6 also improved the granulometric
composition of clinker and decreased dusting. Improved hydraulicity was expected
because of enhanced activity of ferrite and alite modification in the clinker. However,
cement pastes prepared from these clinkers did not show any significant change in the
setting or strength properties when compared to the regular ones (Kakali et al., 1989).
Tungsten is a very high temperature melting metal (melting point=3410°C). Since it will
not volatilize at kiln temperatures, its presence in the CKD, or stacks emissions, is
exceedingly remote.
Manganese
Manganese (Mn) in clinker comes from both the primary and auxiliary raw feeds.
Limestone can contain up to 1.91% Mn2O3 as the carbonate mineral rhodochrosite,
whereas shales and bauxite can have up to 0.59% and 0.37% by weight respectively
(Bucchi, 1981). In blast furnace slags, Mn2O3 can be present up to 1.2% and in coal fly
ash up to 1.44% by weight.
Cement produced from slags can contain more than 1% Mn 2O3 and usually imparts a
brown color to cement (Lea, 1971). The polymorphism of silicate in clinker is affected by
the presence of manganese oxides in the raw material. Knöfel et al. (1984) reported that
the limit of Mn2O3 substitution in C3S is approximately 2.2% at 1550°C. At lower
concentrations, say 0.1% Mn2O3, single substitution of Si+4 by Mn+4 takes place,
whereas at 2.2% Mn2O3 concentration, a double substitution of Si+4 by Mn+4 and Ca+2 by
Mn+2 is possible. The stabilized C3S polymorph was identified as monoclinic; Gutt and
Osborne (1969) reported it to be triclinic. Miller (1976) demonstrated that at low
concentrations (<0.7%), Mn stabilizes monoclinic alite, but at high concentration and in
the presence of fluoride, triagonal alite with markedly high hydraulicity is formed.
Manganese can occur in a number of oxidation states depending upon the burning
conditions in the kiln and can impart different color in clinkers, ranging from reddish-
brown to blue. Puertas et al. (1988) have studied the influence of kiln atmosphere on Mn
solid solutions in C3S and C2S. It was reported that under reducing conditions
isomorphous replacement of Ca+2 by Mn+2 occurs, while in air having higher oxygen
level, Mn+4 replaces Si+4.
According to Knöfel et al. (1983) alite content of clinker increases with Mn addition, with
maximum alite attained at 0.5% MnO 2 and 1% Mn2O3. High Mn content promotes the
formation of belite in the silicate phases, but is more preferentially incorporated into
ferrite phase through the formation of “alumino-manganite”, such as C 4AMn (see Figure
9). This reduces C3A and marginally increases free lime, thus reducing the early
compressive strength of the pastes.
Manganese will not volatilize at kiln temperature (boiling point=1960°C), and is unlikely
to concentrate in the CKD or be found in stack emissions.
Cobalt
Cobalt (Co) is present in traces in the raw mix; the maximum reported concentration is
23 ppm CoO. It has also bee found at much higher levels (up to 1.27%) in some of the
coal fly ashes that could be used as a partial cement raw feed (Bucchi, 1981).
The bulk of cobalt that is present in the raw mix is incorporated in clinker. The CoO level
reported in cement is <130 ppm, but the amounts detected in the alite and belite phase
are only in traces as the bulk of cobalt is concentrated in the ferrite phase by replacing
Fe+3 and forming the “C4ACo” phase (see Figure 9). Cobalt can also give color to
cement.
Sychev et al. (1964) demonstrated that Co somewhat reduces the hydraulic activity of
alite and increases clinker hardness. According to Miller (1976), cobalt increases the
water demand and marginally reduces the late strength of cement paste.
Cobalt is unlikely to vaporize in the kiln (boiling point=2870° C), thus, concentrations in
CKD or in stack gases are expected to be exceedingly small.
Nickel
Sprung (1985) has reported traces of nickel (Ni) in limestone (1.5-7.5 ppm), clay or shale
(61-71 ppm), coal (20-80 ppm), used oil (3-30 ppm), and petroleum coke (208 ppm). In
coal fly ashes, NiO is present up to 1.9% (Bucchi, 1981).
Nickel preferentially concentrates in the ferrite phase, followed by alite, aluminate, and
belite as shown in Figure 9 (Hornain, 1971). Between 0.5 to 1.0% nickel stabilizes alite
(Rangaro, 1977). NiO substitutes for CaO up to 4 mole % in alite and stabilizes the
monoclinic form (Enculeseu, 1974). This alite modification apparently enhances the 1-
day and 5-year compressive strength. Miller (1976) reported that water soluble nickel
compounds act as accelerators and tend to give high early strengths. Kantro (1975) and
Zamorani et al. (1989) also found NiCl 2 to be an accelerator for C3S pastes when used
as mix solution.
Mostly, Ni compounds are non-volatile, yet, owing to the volatile nature of some
compounds, such as NiCO3, nickel could en up in the kiln dust, although recent PCA
studies has shown a maximum of only 60 mg/kg Ni in the CKD. The average amount of
Ni in cement is 31 mg/kg (PCA, 1992).
Copper
Bucchi (1981) has quoted an average of 16 ppm copper oxide (CuO) in the raw mixes,
and a <0.13% in coal fly ash.
Soluble copper salts are retarders and give low heat of hydration (Takahashi et al., 1973;
Miller, 1976; Tashiro et al., 1977). The effect is more pronounced on the C 3A phase
(Tashiro et al., 1979). The addition of copper also gives low sulfate expansion in certain
cases (Miller, 1976).
Silver
Silver (Ag) is present only in traces (<0.250 ppm) in both the kiln raw mix material and
coal; in coal it may occur as silver sulfides or as a complex.
Since silver occurs in traces, it is not expected to significantly contribute in the clinkering
process. Silver is present at 9.2 ppm in cement; it is reported in CKD at 6 ppm for kiln
operated with conventional fuels and at 2.5 ppm for kilns using waste fuels (PCA, 1992).
Zinc
Zinc (Zn) is a trace element in the raw mix, reporting 22-24 ppm in limestone, 59-115
ppm in clay or shale, and 16-220 ppm in coal. However, it can be present up to 3,000
ppm in used oil as a potential secondary fuel (see Table 6), or 10,000 ppm in used tires.
Its concentration is also reported to be significant in alternative raw materials such as
certain metallurgical slags, basic oxygen furnace (B.O.F.), dust, and B.O.F. filter cake
(Miller, 1976, 1994).
Between 80-90% ZnO in the raw mix typically becomes incorporated in clinker (Sprung
et al., 1978; Knöfel, 1978). Approximately half of the zinc is distributed in silicates with
preference for alite while reducing belite; the other half is distributed into the matrix with
preference for the ferrite phase (Knöfel, 1978; Tsuboi et al., 1972). According to Hornain
(1971) zinc in clinker is preferentially retained in ferrite followed by alite, aluminate, and
belite (Figure 9). ZnO additions accelerate the clinker formation. Alite and C 2(AF)
formations increase at the expense of belite and C 3A due to ZnO doping (Odler et al.,
1980-3).
Stevula and Petrovic (1981) prepared a triclinic modification of C 3S of the Type TI-TII
from mixtures of 0.75-1.5% ZnO and pure C3S, fired at about 1600°C and slowly cooled.
Additions of 3.0 and 4.5% ZnO formed rhombohedral C3S with no free ZnO. Boikova
(1986) reported the change in C3S symmetry from triclinic to monoclinic and to
rhombohedral with increasing ZnO additions.
Up to 1.0% ZnO in the raw mix decreases free lime considerably (Odler et al., 1980-1,2),
retards the hydration, and reduces strength when added in excess of 1.0% (Odler et al.,
1980-3). Similar observations were reported by Knöfel (1978). Miller (1976) suggested
the likelihood of reducing Zn in clinker by preferentially vaporizing it where the liquid
phase is low, thus reducing the potentially deleterious effects on cement setting.
ZnO as an admixture also imparts a severe retarding effect on cement hydration; early
strength is reduced, and the late strengths (28-days and beyond) are increased. In fact,
zinc also increases the late strengths (5-10 years) but decreases the paste shrinkage
during the early ages of 1 and 28 days (Miller, 1976). Arliguie et al. (1982, 1985, 1990)
demonstrated that C3S, C3A, and cement hydration are delayed by the formation of
primary zinc hydroxide and its conversion to a crystalline form around the anhydrous
grains. Miller (1976) reported the formation of a complex calcium hydroxo-zincate
intermediate compound that inhibits C3S hydration.
According to Sprung et al. (1978), the volatility of zinc for preheater kiln could be 10-
20%. For a multistage preheater kiln, the capture of ZnO would be more effective and
could result in the total incorporation into the clinker. An average of 149 ppm zinc has
been reported in the CKD from the U.S. plants using conventional fuel and 150 ppm for
those using hazardous waste fuel (CRI, in Mantus, 1992); the corresponding zinc levels
in the plant emissions are only 2.97 and 1.53 mg/sec, respectively.
Cadmium
Cadmium (Cd) occurs in traces in the raw materials and fuels. The average amount of
Cd in cement has been reported to be 0.34 mg/kg (PCA, 1992).
Cadmium in the raw feed reacts with the constituent of kiln gas and can form halides or
sulfates, both are readily vaporized at peak kiln temperature (Kirchner, 1985). The form
of cadmium incorporated in clinker is not known; however; with increasing chloride input
in the kiln, the concentration of Cd in clinker is known to decrease. The addition of CdCl 2
in the raw mix has the same effect. In a cyclone preheater kiln, 74-88% of the total Cd
entering the kiln is incorporated in the clinker as opposed to 25-64% for that produced in
the grate preheater kilns; the remaining Cd is captured in the kiln dust (Weisweiler et al.,
1987).
Recent studies (Bhatty et al., 1993) have shown that high CdO concentrations retard the
cement hydration, but the strength properties are not affected. Addition of soluble
cadmium salts (CdCl2) has no apparent effect on cement hydration. Cd is not leached
from the cement pastes when used as CdO and CdCl2 admixtures.
Mercury
Mercury (Hg) is a trace element. It is highly volatile and vaporizes at much lower
temperatures (boiling point=557°C).
Mercury is somewhat inert, and very little is known on its interaction in the clinker making
process. It is very likely that mercury and its compounds would volatilize in the pre-
calcination region at temperatures closer to 400°C and escape with the stack gasses.
Total mercury is well below detection limits for most North American cements.
Recent studies (Bhatty et al., 1993) have shown that mercury compounds (both
insoluble HgO and soluble HgCl2) impart little effect on the paste hydration and strength
properties.
Elements 51-71 are commonly known as “the rare earths” or “lanthanides”. They are
present only as traces in raw materials and cement clinkers. Owing to their presence at
extremely low levels, they have not been a subject of extensive studies in cement
manufacturing.
Since the rare earths have very similar properties to one another, it is assumed that they
all will have somewhat similar effects on the clinker formation.
Boikova et al. (1964, 1966, 1986) and Toropov et al. (1963) observed the substitution of
rare earths for Ca in both C3S and C2S. The solid solution of C3S with oxyorthosilicates
of lanthanum (La) and Scandium (Sc) results from the similarities in ionic size, and
chemical properties between Ca, La, and Sc. Formation of C 3S solid solution with
gadolinium (Gd), neodymium (Nd) and Erbium (Er) have also been reported. Jantzen et
al. (1979) reported that nearly 15% of Nd 4.Si3O12 can be dissolved into -C2S, which
have identical hydraulicity to that of regular -C2S. La stabilizes all modifications of C3S
solid solutions (Stevula et al., 1981; and Sinclair et al., 1984). Gd gives the triclinic and
monoclinic formulations, whereas Sc produces only the triclinic C 3S solid solutions.
Leaching of Nd stabilized in cement pastes is also fairly low.
Lanthanum also stabilizes the solid solution of C 2S by substituting for Ca+2 (Toropov et
al., 1962-1,-2). Based on the observations on the rare earths, La, Nd, Gd, and Sc,
Boikova (1986) assumed that the remaining rare earths, for having similar chemical and
ionic characteristics, would isomorphously substitute for Ca +2 in C3S and C2S. As a
result, a larger distribution of rare earths in wastes could be expected in the clinker
silicate phases.
Rumyanstev et al. (1970) reported that the addition of La2O3 in the raw mix accelerates
the formation of clinker under laboratory conditions. The engineering properties of the
resulting cements were also enhanced when compared to control. LaCl 3 accelerates the
C3S hydration when added as admixture (Kantro, 1975).
The idea of studying rare earths in cement manufacturing also stemmed from the
possibility of using various medium to low level radioactive wastes that frequently
contained significant amounts of rare earth elements.
Studies by Jantzen et al. (1982), and Boikova (1986) indicated that 20-30% loading of
radioactive waste composed of La2O3, UO3, CeO2, and other oxides, give optimum
elemental retention in clinker at processing temperatures of 1100°C-1200°C.
Volatilization and activation of radionuclides occurred above 1200°C, whereas clinkering
at 1000°C produced incomplete reaction.
Rare earths are expected to have low volatilities (Klein et al., 1975) and are very unlikely
to be found in the kiln dust or in stack emissions.
CONCLUSIONS
The effects of almost all the stable minor and trace elements on the production and
performance of Portland cement have been reported. Emphasis has been given to
elements which occur in natural and by product materials used for cement
manufacturing. The elements for which detailed information has been obtained are dealt
with in an order based on the periodic classification of the elements. The volatilities of
the elements have also been discussed where ever necessary. Elements reviewed
include hydrogen, sodium, potassium, lithium, rubidium, cesium, barium, beryllium,
strontium, magnesium, boron, gallium, indium, thallium, carbon, germanium, tin, lead,
nitrogen, phosphorous, arsenic, antimony, bismuth, oxygen, sulfur, selenium, tellurium,
fluorine, chlorine, bromine, iodine, helium, neon, argon, krypton, xenon, yttrium, titanium,
zirconium, vanadium, niobium, tantalum, chromium, molybdenum, tungsten, manganese,
cobalt, nickel, copper, silver, zinc, cadmium, mercury, and the lanthanides.
The review indicates that although certain elements have been studied in detail for their
role in the clinkering process and cement properties, data on a number of elements is
available only to a limited extent. Although some progress has been made in
understanding the role of these elements on clinkering properties, interaction between
their nature and the major clinker components still needs to be properly understood.
Use of trace elements as fluxes or mineralizers to enhance the clinkering process is also
being realized, yet, understanding of the underlying physico-chemical mechanism and
the potential energy saving aspect requires additional input.
Now that the technological advances in cement manufacturing are in place, efforts can
be directed towards exploring the following aspects of clinker-element interaction:
1. Make use of the minor elements present in raw materials and fuels to enhanced the
clinkering process and the performance of cement.
2. Make use of the minor elements in conserving energy during clinker production;
elements with proven fluxing/mineralizing characteristics could be prime examples.
ACKNOWLEDGMENTS
This report (PCA R&D Serial No. 1990) was prepared by Construction Technology
Laboratories, Inc. (CTL) with the sponsorship of the Portland Cement Association (PCA
Project Index No. 93-01). The author wishes to acknowledge the contributions of W.A.
Klemm and F.M. Miller of CTL for carefully reviewing the manuscript. The contents of the
report reflect the views of the author who is responsible for the facts and accuracy of the
data presented. The contents do not necessarily reflect the views of the Portland
Cement Association.
Titanium Goes in ferrite, decomposes alite to belite, reduces Slows initial setting, improves
melt temperature, gives buff-color cement strength
Zirconium Modifies alite and belite crystals, imparts color Retarder, low early strength
Vanadium Goes into alite, forms larger crystals, reduces melt Increases hydraulicity, causes
viscosity, free lime, effects grindability, lining sulfate expansion
Niobium Feeble effect Little effect
Chromium Reduces melt viscosity, goes largely to belite, Increases early strength, causes
improves grindability, imparts color sulfate expansion
Molybdenum Reduces melt viscosity, forms large roundish alite No adverse effects
crystals, modifies belite crystals
Tungsten Reduces melt viscosity, forms large roundish alite No significant effect
crystals, forms type III belite crystals
Manganese Goes to ferrite, can substitute both Si and Ca in Reduces early strength
C3S, imparts reddish brown to blue color
Cobalt Goes to ferrite, replaces Fe in ferrite, imparts color, Increases water demand, reduces
increases hardness hydraulicity and strength
Nickel Goes to ferrite, replaces Ca in alite and stabilizes Enhances strength, soluble salts
monoclinic form, volatile, reports in CKD function as accelerators
Copper Goes to ferrite, can adversely affect alite and belite Soluble salts as retardes, reduces
formation, lowers melt temperature, free-lime sulfate expansion
Silver In traces, no known effect No known effect
Zinc Enters belite and alite, modifies alite crystals, Oxide as admixture is retarder,
reduces free-lime, improves clinkering low early but high late strength
Cadmium Forms volatile halides/sulfates, enters CKD, Oxide as admixture is retarder
reduces melt temperature, improves burnability
Mercury Somewhat inert, volatile, goes in stack gases Little effect
Lanthanides
Scandium Replaces Ca in C3S and C2S, forms solid solution No known effect
with C3S of triclinic nature
Lanthanum Replaces Ca in C3S and C2S, forms solid solution Accelerates alite hydration
with C3S, enhances clinkering
Cesium Gets uniformly distributed in clinker, have very little Accelerates alite hydration
volatilization
Neodymium Forms solid solutions with C3S and C2S, replaces Improves alite hydration, low
Ca in C3S and C2S leaching