Energy Changes When Kaolin Minerals Are Heated
Energy Changes When Kaolin Minerals Are Heated
A R E HEA TED
By F. VAUGHAN
British Ceramic Research Association, Stoke on Trent.
[Read 6th November, 1954]
ABSTRACT
The work outlined in this paper was undertaken to show how, by a
consideration of the energy changes involved, some insight may be
gained into the reactions occurring when kaolin minerals are heated.
Activation energies have been determined experimentally for the
dehydration of various kaolin minerals. The fireclay mineral is dis-
tinctive and it is suggested that it does not form part of a halloysite-
kaolinite series. It is further thought possible that there is a series of
fireclay minerals.
From thermochemical and thermodynamic considerations the heat of
formation of kaolinite has been calculated approximately and enthalpy
and free energy changes have been estimated for the exothermic reaction
at 980~ These indicate that, if a compound is produced during de-
hydration, the most probable explanation of the exothermic reaction is
the simultaneous production of mullite, ),A1203 and SiO2. Finally
the lattice energies of kaolinite and its thermal decomposition products
have been calculated.
INTRODUCTION
An extensive literature testifies to the considerable research that
has been carried out on the thermal decomposition of kaolinite and
related minerals. Nevertheless a complete picture has not yet been
obtained of the various changes that occur on heating.
(A) Dehydration. In less well-crystallized kaolin minerals sorbed
water is removed in an initial endothermic reaction at about 100~
All minerals in the kaolin group undergo dehydration within the
temperature range 400-700~ but the dehydration temperatures
differ for the various members, being highest for dickite and nacrite.
The start of dehydration is also determined by other factors. The
precise nature of the dehydration product is still not settled but most
workers consider it to be either a poorly-organized compound,
A12Si20 v, or a mixture of alumina and silica. A minority think it to
be similar to a solid solution, and various other theories, not now
seriously held, have been propounded. The literature has been
reviewed by Pieters (1928), Mellor (1938) and others. The de-
hydrated material is essentially amorphous, but pseudomorphs after
the parent mineral are common; a few may persist to very high
temperatures (1,300-1,400~ although other reactions must have
occurred within them.
The researches of both Nelson and Hendricks (1944) and Brown
and Gregg (1952) suggest that the loss of combined water results in a
smaller increase of surface area than might be expected. In this
connection, it is interesting to note that Slawson and Vaughan (1954),
using the electron microscope, have observed that, in the absence of
deflocculating agents, there is an aggregation of the pseudomorphs
265
266 v. VAUGHAN
after kaolinite and dickite. Suspensions of the dehydrated material
were also found to settle more rapidly than unfired kaolinite and
dickite.
(B) High-temperature reactions. Between 900~ and 1,000~ a
marked exothermic reaction occurs, possibly immediately preceded
by a small endothermic reaction, and a second evolution of heat is
observed at about 1,200~ There is no universal agreement about
these reactions and there exists little information about reactions
above 1,300~ The temperatures at which the reactions begin
vary for the different kaolin minerals; they are functions of time and
temperature (Richardson, 1951), and there are other complicating
factors to consider such as catalysis and grain size.
The work outlined in this paper was undertaken to show how
some insight may be gained into the reactions by a consideration of
the energy changes involved. More accurate data than are at present
available may well yield a final elucidation.
KINETICS
Before 1932 there existed little quantitative data on th~ decom-
position rates of clays. Schwartz and Trageser (1932), Budnikov and
Bobrovnik (1936), and Murray and White (1949), obtained informa-
tion indicating that kaolinite dehydrated according to a first-order
law. The last workers estimated the activation energy by using the
Arrhenius (1889) equation, and applied the data to differential thermal
analysis.
(,4) Determination of clay mineral content. The author has studied
a number of cla'ys; china clay, ball clay,fireclay, dickite and halloysite.
Chemical analyses, petrological, electron-microscope and X-ray
studies, d.t.a, examinations and estimations of organic matter and
carbonate content were made. Halloysite and dickite contained no
impurities capable of interfering in the dehydration studies. The
china clays, ball clays and fireclays all contained some mica but, at
the dehydration temperatures used, its interference is negligible. The
organic matter in the ball clays and fireclays was removed by hydrogen
peroxide treatment; an initial heating (below 400~ of halloysite,
ball clays and fireclays ensured the removal of sorbed water. Re-
producible dehydration curves were best obtained after the ball-,
china and fire-clays had been ground to pass a 120 mesh B.S. sieve,
but the dickite and halloysite were used as received.
For each mineral a weight loss-v-tifne curve was determined by
placing weighed samples in a small laboratory electric furnace main-
tained at a constant temperature. Temperatures from 450 to 550~
proved adequate in studying all the minerals except dickite for which
temperatures between 550~ and 650~ were necessary.
Two methods were used to estimate the amount of clay mineral
initially present, that o f Murray and White (1949) and a method
based on the following equations:
k = l [loge xl - loge (X2 - (1)
ENERGY CHANGES OF HEATED KAOLINS 267
and
X 2 --X 1
a ~- (2)
e - kt 1-e - kt2
where k is the velocity constant at a particular temperature in the
dehydration range ( e . g . , 550~ a is the amount of clay mineral
initially present, and x t , x 2 are the amounts decomposed in times
t 1, t z respectively. The full derivation o f these equations will not
be given here but equation (1) is only valid when t 2 = 2 t r The weight
losses occurring at 10 and 5 minutes, 12 and 6 minutes and so on,
were derived from the loss-time curves and, by substitution in equation
(1), k was calculated. Both methods confirmed that the dehydration
was a first-order reaction, and for such reactions k can be substituted
in equation (2) and the amount of clay mineral (a) calculated.
Values of a obtained by the Murray and White method agreed quite
well with those obtained from equations (1) and (2); they should, of
course, yield identical values since both methods are based on a
first-order equation. The differences between them may be a
measure of the accuracy of the experimental dehydration curve.
(Owing to the small amounts of dickite and halloysite available full
dehydration curves for these minerals were not obtained, and a
was estimated on the basis of the preliminary studies mentioned
earlier). The results are given in Table 1 where the clay contents
obtained using the author's procedure are compared with those
obtained when using the Murray and White method.
TABLE 1--Clay mineral contents of clays using different methods.