CEMENT and CONCRETE RESEARCH, Vol. 20, pp. 862-868, 1990. Prinmd in the USA.

0008-8846/90. $3.00-~OO. Copyright (c) 1990 Pergamon Press plc.

FLUXING REACTIONS OF SULPHATE AND CARBONATES

[N CEMENT CLINKERING. I. SYSTEMS
CaSO4-K2SO4 AND K2SO4-CaCO3

Hugo B o l i o Arceo~ and F.P. Glasser

Department of Chemistry, U n i v e r s i t y of Aberdeen,
Meston B u i l d i n g , Meston Walk, Old Aberdeen, AB9 2UE, Scotland.

Current Address: Cementos MexJcanos. S.A., I n v e s t i g a c i o n y D e s a r r o l l o ,

Ave C o n s t i t u c i o n 444 PTE, APDO Postal 31GO, Monterrey, Nuevo Leon, Mexico.
Mexico.

(Communicated by J.P. Skalny)

(Received March 30, 1990)

ABSTRACT

Phase e q u i l i b r E a and thermodynamics h a v e b e e n used to determine

c o m p a t a b i l i t i e s in the systems CaSO4-K2SO4 and K2SO4-CaCO3. In the
CaSO¢-K2SO4 system, a polymorphic inversion at 200"C is encountered
in K2Ca2(SO4)3. Thermodynamic calculation of the reciprocal salt
system K2SO4-CaCO3-CaSO4-KzC03 discloses that the join K2SO4-CaCO3
constitutes a stable binary: experiments in flowing CO2 at ! bar
pressure confirm this. A noteworthy feature of the system is the
solid solution of CaCO3 in K2SO4, reaching ~20% mol % CaCO3. The
eutectic, located at 880"C and 43 mo] % CaCO3, is attainable without
significant decomposition. The possible consequences to clinkering
reactions are discussed.

INTRODUCTION
The batching of cement-making raw m a t e r i a l s is g e n e r a l l y p r o p o r t i o n e d so
as to include f l u x e s : i r o n and aluminium oxide as well as o t h e r , less
obvious f l u x e s such as sulphate and carbonate. A 1 2 0 3 and Fe203 are most
e f f e c t i v e at high temperatures, but the other f l u x i n g agents are p o t e n t i a l l y
more a c t i v e at lower temperatures. The tendency to separate the o v e r a l l
c l i n k e r i n g process i n t o two stages allows new o p p o r t u n i t i e s to employ f l u x e s
Jn the lower-temperature stage as, f o r example, in the suspension preheater.
However, the sequence of events o c c u r r i n g at the lower temperature stage i s
less well documented than the main c l i n k e r i n g r e a c t i o n s . This paper
presents new data on the r o l e of sulphates and carbonates.

EXPERIMENTAL
Batches intended for phase equilibrium studies were prepared from
ana]ytlca~ reagent grade materials on a ca I 0 gram scale. The r e a c t a n t s

862
Vol. 20. No. 6 SULFATE, CARBONATE. FLUXING.CLINKER 863

were blended by hand in agate mortars and stored in dry conditions u n t i l use.
The bulk compositions of representative chan~es w e r e checked by chemical
analyses f o r t h e i r p r i n c i p a l components.
The systems studied do not generally quench to yield glasses. However,
standard quenching techniques could still be used; ca 50-100 mg samples.
contained in gold or platinum foil, were suspended in vertical tube,
electrically-heated quench furnaces controlled to ~2"C. Temperatures were
read against either chromel-alumel thermoelements (below 1000"C) or, at
higher temperatures, by Pt - 13g Rh thermoelements. The condition of the
quench runs could generally be established by petrographic examination: with
practice, devitrified melt was readily distinguishable from primary crystals.
The temperature and position of supposed invariant points was also determined
by differential thermal analysis (DTA) using a Stanton-Redcroft STA 781
instrument. This has the facility of simultaneous thermogravimetric
analysis. Thus, not only could invariant temperatures be determined, but
the possibility of compositions drifting, owing to vapour-phase losses, could
also be monitored continuously. In general, determinations were deemed
invalid when the cumulative weight loss of charges, other than from sorbed
moisture, exceeded -1%. DTA was also useful to monitor rapid, non-
quenchable solid phase transformations.
Phases present were identified optically, with a petrographic
microscope, as well as by X-ray powder diffraction using CuK~ radiation in
conjunction with either a Philips diffractometer or H~gg-Ouinier camera.

RESULTS AND DISCUSSION

System CaSO4 - K~SOd. This system has been studied by a number of
investigators(l-4). In the course of the present study, the maximum extent
of solid solution of CaSO4 in K2SO4, the position and temperature of the
eutectic, and the incongruent melting of K2Caz(SOd)3 KC2Sa were
redetermined. In addition, a polymorphic transition in KC2S3 was detected
by DTA at I00" ±5"C. Table I records some of the data obtained in the course
of the present study and compares it with data obtained previously;

Table I C r i t i c a l Features of the CaSO4 -K2SO4 System

Reference
Feature Muller(2) Bellanca(3) Rowe(l) Pliego(4) Present Study
K2SOd(m.p.) 1057 1071 106g -
CaSO4 (m.p.) 1450 1380 1462
Eu~ectic
temperature 867 876 875 869 878
wt % CaS04 36"2 34.3 34'0 34"i 34.2
Polymorphic transformation
High-low,K2~34 554 535 540 540 557
Maximum S o l i d S o l u t i o n of CaSO4 in K2SO4
wt % K2SO4 14.6 19-8 20 18 20±4

Temperatures are "C: m.p. = melting p o i n t .

the amended phase diagram is shown i n Fig. i . Generally good agreement is

obtained between the r e s u l t s of d i f f e r e n t studies and the r e s u l t i n g diagram
d i f f e r s l i t t l e from that presented by Rowe(1): note however the data which
have been added concerning the poIymorphic t r a n s i t i o n in KC2S3.
Some doubt remains about the polymorphism of KCa(SOd)2. Three
polymorphs h a v e been reported. The y phase, defined by DTA evidence and
864 H,B. Azcco and F.P.Glass~ Vol.20, No, 6

14620

1195 °

1100 L • c~
1069" - KZ ""2 S 3" L ,
1011 °
KzCz~s • c~
940 a
900 OTO*

700 ~.K2~ , K2Cz~,3

KzCz~'3
~ 511J C5
540" Fig. I Revised Phase Diagram of
~e s00
E the K2SO4-CaSO System
)..
DKZS * K2C2$ 3
3OO
200 °
~BK2~ • K2C2~] K2C2P,3 • C~
100 , i , I , I I
0 20 /,0 6'O ' 80 100
K2 SO~, WEIGHT % CaS0~

defined by i t s crystallographic properties as b i a x i a l , orthorhombic(5),

appears to be s t a b l e below 200"C. The powder X-ray data obtained f o r the
phase s t a b l e above 200"C, designated ~, can be indexed on the basis of a
cubic c e l l : our r e c a l c u l a t i o n of Morey's data gives ao = I0.424 A at 200"C.
The low temperature phase i n v e r s i o n , ~-¥, involves a considerable s p e c i f i c
volume change which we c a l c u l a t e to be 2.5% when c o o l i n g from 2 0 0 " C down to
room temperature and i s accompanied by sample " d u s t i n g " , which could lead
to a c o n t r i b u t i o n to k i l n dust formation in t h i s temperature range.
Morey(6) r e p o r t s that a t r a n s f o r m a t i o n was detected at ~940"C, above which
the ~ phase e x i s t e d : ~ and B ere said to have i d e n t i c a l powder p a t t e r n s .
This i s p o s s i b l e , although i n our experience phase changes associated w i t h
such s l i g h t symmetry changes are u n l i k e l y to give r i s e to the large thermal
anomalies which were observed by DTA. The present authors regard the case
f o r the existence of an ~ polymorph as proven, but conclude that some
u n c e r t a i n t y e x i s t s concerning i t s symmetry and powder data.

The extensive s o l i d s o l u t i o n of CaSO4 in K2SO4 i s noteworthy. Presumably

the mechanism of solid solution involves o n e Ca2+ and a cation vacancy
substituting for two K + ions. Eysel (7) confirmed this mechanism of solid
solution formation when describing the analogous solid solution of CaSO4 in
Na2SO4 (1); this is relevant because the latter is isostructural or nearly
so with high K2SO4.
System K%SO4 - CaCO~. In studies of the CaCO3-CaSO4-K2SO4-K2C03 system,
Luginina et aI(8-10) leave unresolved whether K2SO4 and CaCO3 constitute a
stable binary pair at any temperature. In one study(8) they report that this
pair constitute a thermodynamically stable binary. However, in other
studies the same group(9-10) claim that the reciprocal pair K2CO3-CaSO4 are
instead the more s t a b l e .
At any one pressure and temperature, only one possibility can be
correct In the course of the present study, both thermodynamic calculations
and practical experiments were undertaken.
Standard thermodynamic data were used for the calculation of the
energetics of the reaction:
Vol. 20, No. 6 SULFATE, CARBONATE, FLUXING,CLINKER 865

K2C03 + CaSO4 = CaCO3 + K2SO4 (I)

Table If. Thermodynamic Data for the Reaction

K2CO3 + CaSO4 = CaCO3 + K2SO4
AHzgs" AS29a" A B x 103 C x 10s
Compound Kcal/mol CaIlmol/K
CaC03 - 288.45 22.2 2~--98 5.24 -6.2
CaSO4 - 342.42 25.5 18"52 21.97 -1"57
K2CO3 - 273.93 35.7 19-19 26.06 -
~K2SO4 - 342"66 42.0 28.77 23.80 4"26
~KzSO4 - 33"60 13.40 -

AS transition forK2CO3 and K2SO4 are 700 and 2140 cal mo]-1 respectively, at
695 and 856K respectively.
and these are given in Table II. The data are taken from critical
compilations (II). Basically, the Gibbs free enerEy function, AG, was
solved for the reaction (I), transforming the relevant thermodynamic values
to successively higher temperatures by means of the relation:
ACT = AH298" - TAS298" + ;T298ACp dT - T ; T 2 9 8 A C p dT (2)
T
where T is the temperature in Ke]vins at which the calculation i s made and Cp
the heat capacity at constant pressure: ACp i s g i v e n b y :

ACp = Z C p ( f i a . | ) - ~Cp(initi,l) (3)

where 'final' and 'initial' refer to the different sides of equation(1).
The heat capacity is transformed to other temperatures using the relationship
ACp = A + BT + CT2; coefficients ar given in Table If.
Application of equation(2) reveals that when CaCO3 and KzSO4 are taken as
ceactants, the overall free energy to form K2CO3 and CaSO4 is always positive
by -16-22 K cal mol-l, the exact value depending on temperature: Fig. 2
shows the results of the calculation. The conclusion is clear: i.e.
reaction (I) cannot proceed in the proposed direction.

Despite the rather larse

calculated drivin$ force
2S disclosing that the stable pair
is (K2SO4 and CaCO3) the
results are perhaps not
absolutely conclusive; some
2O uncertainty exists concerning
the numerical accuracy of the
data and, moreover, the
thermodynamic calculations had
to be made assumin~ that (i) the
heat capacity of the low temper-
ature polymorph of K2CO3 was
I I t I applicable to the high
300 SO0 700 856 900 1100 temperature polymorph,(ii) the
TEMPERATURE (K) liquid and vapour phases could
be neglected (melting was later
FCg. 2 C a l c u l a t e d f r e e energy for proven to occur at less than
the r e a c t i o n : 900"C), and (iii) no inter-
CaCO3 + ~2SO4 ~ K2CO3 + CaSO4 mediate compound existed.
866 H.B. AfceoandF.P.Glasser Vol. 20, No.6

Experimental d e t e r m i n a t i o n s were a c c o r d i n g l y undertaken to supplement

the thermodyanmic c a l c u l a t i o n s . Dried r e a c t a n t s were mixed and heated in a
CO2 atmosphere at a t o t a l pressure of I bar, to suppress d i s s o c i a t i o n .

Compositions were prepared at 20, 40, 60 and 80 mol % using two sets
of reactants: one consisting of mixtures of (K2CO3 + CaSO4) and the other,
of ([(zSO4 - CaCO3). Mixtures were heated l-3hr in the temperature
range 700 - 850"C.

1000 • ~o ~'R2bu4 * u ~
(&3 z2%)

/"
i"''/I/'I

Cat03 • L

~
~-K2S04ss ~ ~'~'~=~- - - -o. . . . . . . . . -o. . . . . . . . . . . . .
%) o " " 880 ZS° ~ °

L_J
o 800
i,i
EK
( x K 2 S 0 4 * Ca[O 3

CK 0cK2SO4 ss
i,i
O.
600 ~KzSO4ss
u.J ~.7o) = o o O n
o 560 ± 5 ° o
I--
~.B K2SpL'
P K2S04 • Ca[03

400 I I I I I I I I I
10 20 30 40 50 60 70 80 90 100
K2S04 M O L % CaCO 3 CaCO3
Fig. 3 Phase diagram o f the K2SO4 - CaCO3 system at i bar CO2

Mixtures of {(2C03 and CaSO4 were found to develop K8SO4 and CaCO3
although reaction was often incomplete and, at 850"C, melting occurred in
some samples. On the other hand. mixtures of K2SO4 and CaCO3 persisted
unchanged and unmelted at temperatures between 700 and 850'C: no
significant dissociation was observed. The experimental evidence is,
therefore, in accordance with the thermodynamic data: at I bar pressure and
in the temperature range 700"-850"C, the pair K2SO4 and CaCO3 constitute a
thermodynamically stable assemblage relative to K2C03 and CaSO4.
Twelve compositions were prepared at well-spaced intervals from K2SO4 and
CaCO3 for phase equilibrium investigation. The thermal analyses were done
in flowing C02 atmosphere and simultaneous TGA showed that weight losses were
negligible up to ca 880"C. Indeed, in compositions low in CaCO3,
(45 mo~ %, weight loss became significant only at T >900-920"C. The DTA
experiments were supplemented by microscopic observations on pressed pellets
of the reactants which had been placed in Pt-foiI crucibles heated in CO2 at
a constant temperature and subsequently quenched. Additionally, the
presence of CaCO3 was determined by X-ray powder diffraction. Experience of
calibration samples, made by weighing out mixtures of K2SO4 and CaCO3, showed
that as Iitt!e as i-2 weight X CaCO3 could be reliably detected by powder
diffraction.
Vol. 20, No. 6 SULFATE, CARBONATE. FLUXING, CLINKER 867

The resu!t~ng phase diagram i s shown in Fig. 3. No traces of K2CO3 or

CaSO4 were found to occur under any c o n d i t i o n s . V i r t u a l l y no s o l i d s o l u t i o n
of KzSO4 in CaCO3 occurs, but s o l i d s o l u t i o n of CaCO3 in K2SO4 i s extensive,
reaching a maximum of 20 Z4% at 880"C. Solid solution is however much less
extensive at lower temperatures, reachin~ only 5X at 5 6 0 " C ; nevertheless,
even t h i s extent i s s u f f i c i e n t to depress the temperature of the ~-B K2SO4
inversion from ~84"C to 560" ~5"C.

The e u t e c t i c composition at 43 mol % CaCO3 and a temperature of

880"C, was attained at this CO2 pressure without great weight losses.
However, at only slightly higher temperatures, ca 900-910"C, rapid
decomposition with loss of CO2 occurred. The dashed lines in Fig. 3 show
this: extension of the liquidus towards CaCO3 would require CO2 pressures
>I bar to attain, so the horizontal dashed line represents a temperature-
composition cut-off at the maximum COz pressure used in the course of this
study.

DISCUSSION
The study o~ these fused salt systems reveals some interesting
reactions relevant to the initial stages of cement cllnkering. The
importance of Ca-langbeinite as a reaction product between CaSO4 and R2SO4 is
=on~irmed. However, this phase is known to evaporate incongruently at high
temperatures, liberating K and S species to the vapour phase (I0).
Condensation of these species may occur at some points in the kiln system and
at these points, K2SO4 may also appear. It is shown that if CO2, CaO or
CaCO3 are also present, the resulting K2SO4-type phase need not be
chemically pure: extensive carbonate-containing solid solutions may instead
form. These are most extensive close to the eutectic temperature which,
very nearly, coincides with the decomposition point of CaCO3 in a CO2-rich
atmosphere. This coincidence is fortuitous but significant because both
occur within the normal working range of suspension preheaters.
On rapid cooling, these solid solutions may be preserved to ambient.
On the other hand, with slower cooling, they may exsolve with the result
that CaCO3 will be formed. The only polymorphic phase of CaCO3 detected as
product of this exsolution was ca:cite.
The occurrence of a liquid relatively rich in molten carbonate is
~gnificant. Normal CO2 pressures in cement kilns do not reach I bar but,
as additional components are added so as to further depress fusion
temperatures, carbonate-containing melts may well persist into the range of
temperatures and CO2 pressures and temperatures norma|[y experienced in
suspension preheaters. Molten carbonates are known to be good fluxes for a
wide range of oxides and it is thus possible that the formation of low-
melting point, carbonate-rich melts could significantly influence the course
of assimilation reactions of the main components - CaO, SiO2, etc.
It is known from analyses of ground cement clinker that they often
contain "CO2". The source of this CO2 is generally believed to arise from
reaction between the ground clinker and the atmosphere, as may occur durin~
storage. While not wishing to minimize the importance of this source of
carbonation, it is noteworthy that cements known to contain a K8SO4
condensate may also carry significant CO2 in the form of an alkali sulphate-
carbonate solld solution. The shift in X-ray d-spacings arising in the
course of solid solution formation is quite sma|I, so even at its m a x i m ~
extent its widespread occurrence in clinker may escape detection.
Nevertheless, the CO2 content of a clinker may arise from several sources,
one of which condensation of a sulphate-carbonate phase during clinker
cooling - is highlighted. Redistribution of this carbonate during storage
may be one cause of pack set.
868 H.B.A.rceo and F.P. Glasser Vol. 20, No. 6

ACKNOWLEDGEMENT

One of us (HBA) acknowledges the financial support of Cementos

Mexicanos S.A. for financial support.

REFERENCES

J.J. Rowe, O.W. Morey and I.D, Hansen, The Binary System K2SOd-CaSOd.
JourL !norg. Nuclear Chem. 2_~753-58 (1965).
2. H. MflI!er Neus Jahrbuch f~r Mineralogie. Tail 30, I, (1910).
3. A. Bellanca L'aftitalite ne| sistema temario K2SOd-Na2SOd-CaSOd.
Periodico di Mineralogia 1_~321-8G (1942).
4. Y. Pliego-Cuervo and F.P. Glasser. The Role of Sulphates in Cement
Clinkering Reactions: Phase Formation and Melting in the system CaO-
Ca2SiOd-E2SO4 Cement Concr. Res. 7 4787-482 (1977).
~'lational Bureau of Standards, Mono. 25, sac. 7, 39 (1969), JCPDS
Powder Data File (card 20-867).
6. G.W. Morey, J.J. Rowe and R. Fournier, The System K2Mg2(SOd)3
(Langbeinite)-K2Ca2(SOd)3 (Calcium Langbeinite). Jour. Inorg. Nuclear
Chem. 27 53-58 (!965).
7. Eysel, W.; "Crystal Chemistry and Structure of Na2SOd(I) and its
SolEd Solutions", Acta Cryst., B41, 5-II, (1985).
8. I.G. Luginina, A.N. Luginin and G.P. Polvakov. "Reactions In The Open
System CaSOd'nH20-K2CO3. Khim. k'him Technol. (Alma-Ate) 2 eag-a3a
(1971).
9. I.G. Luginina, A.N. Luginin, M. Shaposhnikova and V.K. Klassen,
Chemical Effects of Additives on the Processes Preceding Clinker
Formation. Natural Volatile Additives. Silikatechnik, 24 374-377
(1973).