Point of view

Why Chemistry
Matters in Concrete
BY VAGN C. JOHANSEN, WALDEMAR A. KLEMM, AND PETER C. TAYLOR

W hy does chemistry matter? It matters because chemistry

controls the life span of concrete.
structures made with Aspdin’s cement still stand today, none
the worse for wear.
Chemistry explains why cement hardens and the inter-action The important point is that concrete (the system of cement
between cement and its environment. We will discuss the paste binding together aggregate particles into an artificial stone
basic inorganic chemistry of cement and concrete under body) is inherently reactive and, given its particular individual
service conditions. Of course, there are other types of structure and exposure conditions, that reactivity will result in
chemical processes that occur when mixing concrete, such as either excellent durability or poor durability. In both the long
arcane aspects of organic chemistry and surface chemistry. and short terms, it is the chemistry that makes the difference in
These are left out, not because these subjects are unimportant, concrete perma-nence. Examples of typical exposure agents that
but because they are not the main focus of this article. affect durability are:

■ Moisture and ground water;

Chemistry and concrete ■ Temperature cycles;
The first principle to understand is that, in a broad sense, ■ Deicing salts;
concrete is thermodynamically unstable. When cement paste is ■ Marine environment; and
exposed to the earth’s atmosphere, it begins to deteriorate, which ■ Carbon dioxide and acidic air pollutants (NOx
is a form of chemical corrosion. The paste will react to exposure and SOx ).
to acid rain or just the normal amount of carbon dioxide in the Depending upon the composition of the concrete and
air. This reaction causes surfaces to etch and carbonate, and the exposure conditions, a variety of possible chemical reactions
then, the calcium silicate hydrates that give concrete its strength may deteriorate concrete. Sometimes, how-ever, the enemy is
will be converted back to calcium carbonate, silica gel, and not some outside element, but rather, the seeds of destruction
alumina gel. may be innocently contained within the concrete itself.

This is one aspect of chemistry, but chemistry is not only Moisture, or water, plays the most important role in the
thermodynamics, it is also kinetics. In other words, concrete has setting, strength development, and eventual deterio-ration of
the potential to change, but how fast will that happen? Concrete concrete. Portland cement hardens because of a chemical process
made carefully with the right materials in the proper proportions, called hydration. This means that the silicate and aluminate
and developing the optimum microstructure by adequate curing, minerals in portland cement react and combine with water to
can last for many hundreds, or even thousands, of years. An produce the “glue” that holds together the aggregate that we call
example of this is the Pantheon in Rome, expertly constructed concrete. Further-more, in a more general fashion, these cements
with a pozzolanic “Roman Cement,” which so far has lasted for are called “hydraulic cements.” Hydraulic simply means the
over 2000 years. capability to harden under water. Portland cement is only one
type of hydraulic cement, and it should be noted that C 1157-00,
When Joseph Aspdin patented his “Portland Cement” in the ASTM performance specifi-cation for cement, is called the
1824, it was titled “An Improvement in the Modes of Producing “Standard Performance Specification for Hydraulic Cement,” and
Artificial Stone.” Concrete was considered to be artificial stone includes both portland and blended types.
having the appearance, strength, and durability of the real thing.
Some of the early concrete
As Duff Abrams emphasized in his seminal paper “Design of
The opinions expressed in this point of view article may Concrete Mixtures,” however, presented at the December 1918
not necessarily be those of the American Concrete meeting of the Portland Cement Associa-tion and then published
Institute. Reader comment is invited. as the Lewis Institute Bulletin 1, the water-cement ratio (w/c)
dictates the strength of

84 MARCH 2002 / Concrete international

concrete. Although the size and grading of the aggregate and the durability. One example of this destructive behavior is the
quantity of cement influence the quantity of water required to limestone obelisk in New York City’s Central Park. It had
produce a workable mixture, the amount of water in a mixture survived thousands of years in the Egyptian desert, remaining in
controls the concrete’s strength. There-fore, one should use the almost pristine condition, but now after only decades of
smallest quantity of water that will produce plastic, or workable, exposure to acidic gases in the city’s atmosphere, the
concrete. Water may, depending upon the quality and nature of hieroglyphic inscriptions have all but disappeared from view.
the aggregate used and the concrete curing history, trigger such Although limestone isn’t concrete, the overall effects of acid
other deleterious events as alkali-silica reactivity (ASR) or rain are quite similar.
delayed ettringite formation (DEF). Water may also function as During the design of a concrete mixture, chemistry must be
considered. For instance, if the concrete is placed where it may
a transport medium for potentially aggressive species, such as be exposed to aggressive environ-ments such as chlorides or
sulfate, to enter the system. These chemical reactions will be sulfates, a different ASTM cement type will be selected. Type I
covered in greater detail further in this article. is ordinary portland cement, whereas Type II cement provides
moderate resistance to sulfate attack, and Type V cement
Temperature affects the rate of chemical reactions, and a provides much greater resistance to sulfates. There are now
general “rule of thumb” is that a chemical reaction rate doubles many ASTM standard specifications, ASTM standard test
for every 10 °C increase in temperature. Thus, temperature methods, ACI guidelines, and other recommendations for the
influences the rate of both concrete setting and hardening. engineering professional on how to design concrete mixtures,
Additionally, cycles in tempera-ture may result in freezing-and- and all of these are based upon fundamental knowledge of the
thawing distress if the concrete lacks sufficient air entrainment. chemistry of cement and concrete (Weaver 1978).
Curing concrete above certain critical temperatures may lead to
the expansion and cracking associated with DEF.
During the past 10 to 20 years, field experience has shown
Several external environmental factors may initiate that not only mixture design, but also the curing of concrete is
destructive chemical reactions in concrete, particularly concrete very important. Curing, of course, is the process that provides
with a more open porosity (due to an elevated w/c). Some of sufficient moisture and thermal energy to promote the hydration
these factors include chloride-containing deicing salts used to process. Curing conditions control strength development and
treat roadways in winter, salt spray or tidal exposure to thermal cracking; therefore, they have a significant impact on
seawater, and sulfate-containing soil or ground water. Chlorides the durability of concrete. A recent example of the importance
can slowly diffuse into concrete and, in the presence of moisture of curing and temperature control is concrete degradation from
and oxygen, will initiate corrosion of reinforcing steel. The DEF. As will be discussed later, ettringite formation is a normal
oxidation of iron to produce iron oxide is a chemical process and useful event as portland cement begins to set. If its
that yields a large volume of oxidation product that not only formation is greatly delayed (days or months after concrete
structurally weakens the metal, but also produces internal hardening), however, it can cause serious durability problems.
localized pressure that can cause severe cracking of the concrete
cover. Once cracking begins, more of the concrete surface is
exposed to further chemical attack.
The chemical reactions of cement
Portland cement contains calcium silicates and calcium
A vulnerable part of the concrete is the cement paste. aluminates formed by a sequence of thermal and chemical
Although concrete is composed of 10 to 15% by mass of processes, including decomposition of limestone; reaction with
portland cement, it becomes the focus of aggressive outside other quarried materials such as clay, iron ore, and sand; partial
chemical agents like atmospheric carbon dioxide (CO2 ) and fusion of these ingredients; and finally, the formation of hard,
acidic gases that dissolve in moisture to produce acid rain. rounded nodules called clinker. All of this occurs at temperatures
Cement paste is highly alkaline, with a pH greater than 12.5. reaching about 1450 °C in a cement plant’s rotary kiln. After
This high pH is due to the presence of the calcium hydroxide cooling, the clinker is ground together with approximately 5%
hydration products and the lesser amounts of alkali (sodium and gypsum (calcium sulfate dihydrate) to a flourlike fineness,
potassium) salts. Under ideal carbonation conditions (50 to producing the final product, portland cement (Kosmatka and
70% relative humidity and an exposed paste surface), the Panarese 1994).
hydrated lime (calcium hydroxide) constit-uent reacts with CO 2
to form calcium carbonate, which is the same mineral as calcite At this stage, we will review some elementary cement
or limestone. As this process slowly progresses, perhaps even at chemistry. The present knowledge of the chemical composition
rates of only a millimeter or less per year, the pH is gradually of portland cement and what happens to it when it is mixed with
lowered and a finely crystalline calcium carbonate replaces the water was first disclosed in 1887 by the French chemist Henry
hydration products. Le Chatelier (1905). In his doctoral thesis, he correctly identified
the major cement minerals as tricalcium silicate, dicalcium
silicate, and tricalcium aluminate. In 1915, scientists at the Geo-
Acid rain is aggressive and can more rapidly etch and physical Laboratory in Washington, D.C., were studying the
corrode exposed surfaces, eventually destroying the hydrated high-temperature phase relationships of the ternary
cement minerals that provide strength and

Concrete international / MARCH 2002 85

system CaO-SiO2-Al 2O 3. Among the mineral phases stable calcium sulfoaluminate compound called the
investigated were, of course, tricalcium silicate, dicalcium “monosulfate,” C3 A·Cþ H12 (Eq. (1) and (2)). A fourth major
silicate, and tricalcium aluminate. In the process of publishing mineral in cement, the iron-containing ferrite phase, or
the complex triangle-shaped phase diagram, certain tetracalcium aluminoferrite (C4 AF) also hydrates, although
simplifying abbreviations for the chemical compositions of much more slowly, to form chemically similar trisulfate and
each mineral phase were invented. For example, tricalcium monosulfate compounds, in which iron (Fe2O3 ) replaces a
silicate, Ca3 SiO5, could also be written as the combined portion of the aluminum (Al2O3 ) (Steinour, 1958). Cement
sequence of the two oxides, such as 3CaO·SiO2. The chemists generically call these the AFt (aluminate-ferrite-
investigators, Rankin and Wright, then used the shorthand trisubstituted) and AFm (aluminate-ferrite-monosubstituted)
notation of CaO = C; SiO2 = S; and Al2O 3 = A. Accordingly, phases, respectively.
3CaO·SiO2 could be written as C3S. In similar fashion,
C3 A + 3Cþ H2 + 26H → C6 Aþ 3 H32 (ettringite) (1)
dicalcium silicate, 2CaO·SiO2, became C2S, and tricalcium
aluminate, 3CaO·Al2O3, became C3 A (Bogue and Steinour 2C3 A + C6 Aþ 3 H32 + 4H → 3C4 Aþ H12
1961). (monosulfoaluminate) (2)
This notation was so convenient and useful that, in future
publications over the following years, a shorthand notation for The major strength development of concrete, how-ever,
other oxides was similarly introduced. Thus, results from the hydration of the calcium silicate phases (C3S
Fe2O3 = F; MgO = M; H2O = H; Na2O = N; and K2O = K. The and C2S). Both of these calcium silicates combine with water
additional “S” problem of sulfur trioxide, SO3, was simply to form the gel-like calcium silicate hydrate, or C-S-H (Eq. (3)
resolved by indicating it as þ (S-bar). Other names for cement and (4)).
minerals or their hydrates concurrently entered the language and 2C3 S + 6H → C-S-H + 3CH (3)
are commonly used today. For approxi-mately 100 years, the
tricalcium silicate and dicalcium silicate minerals, in their 2C2 S + 4H → C-S-H + CH (4)
somewhat impure composition as they crystallize in clinker, have
Cement paste microstructure
been called “alite” and “belite,” respectively. Others are called by
their equivalent mineralogical name, such as “periclase” for MgO When concrete hardens due to hydration of the cement paste
crystals in cement, “portlandite” for calcium hydroxide (CH in portion, the total volume of the hydration products is smaller than
shorthand notation), and “ettringite” for the calcium the original volume of portland cement and water. As a result of
these hydration reactions and the decrease in paste volume, a
aluminosilicate hydrate (C3A·3Cþ·32H). The formula for
system of capillary pores is created (Fig. 1). This porosity in
ettringite in cement chemist’s notation appears to be complicated, concrete governs its degree of susceptibility to various chemical
but it is simplified compared to a conventional chemical formula, factors having a profound influence on its durability.
{Ca6[Al(OH)6]2·24H 2O}·(3SO4 )·(2H2O).
The reactions of greatest interest in the cement paste are those
Normal hydration reactions between the components of the pore solution, and the solid
The most rapid reaction that occurs when mixing cement and phases present, which are the original cement minerals, their
water is the hydration of tricalcium aluminate (C3 A). Entirely by hydration products and, possibly, suscep-tible aggregate surfaces,
itself, C3A and water will quickly form calcium aluminate in the case of ASR. This means that, as time passes, some of the
hydrates such as C4 AH13 and C2 AH8. This can occur so rapidly solid phases dissolve and new phases precipitate. The transport in
that the concrete may stiffen within minutes and become entirely solution of the various substances participating in the reactions
unworkable because of the heat emitted. This condition is called controls the rate at which the chemical reactions take place.
“flash set.” In the nineteenth century, when cement developed
strength slowly because it was coarsely ground and imperfectly In order for water to be available, it has to be able to enter the
reacted, flash set was not a problem. Even-tually, with the concrete and move through the paste structure, where it becomes
introduction of rotary kilns, a more scientific proportioning of part of the pore solution within the capillary pore system. The
raw material ingredients, and much finer clinker grinding, the pore solution contains various chemical compounds that have
dissolved from solid materials. These are primarily hydroxyl ions
addition of gypsum (CþH2 ) to the cement eliminated flash- + +
setting problems. (OH–) and alkali ions (K and Na ), as well as lesser amounts of
calcium, silicate, aluminate, and sulfate ions. Ions from external
Chemically the C3 A, the gypsum, and the water would form sources, such as deicing salts, may include sodium and chloride.
a protective coating of calcium sulfoaluminate hydrate This is the point where the impor-tance of the paste
(ettringite) over the exposed tricalcium aluminate surfaces that microstructure, in the form of the capillary pore system, is
would remain for several hours. The C3 A hydration reactivates evident. If cracks in the aggre-gate are disregarded, only the
as the initial setting begins, consuming the sulfate and forming paste will take part in the transport of reactive ions into a
more ettringite. If the portland cement is an ASTM Type I, concrete system. In such a system, there are three possibilities for
however, it probably contains over 8% C3 A, which is a much water movement:
greater percentage than the sulfate present. When all of the ■ Through the capillary system;
sulfate has been combined into ettringite, the excess C3 A ■ Through the hydration products (primarily C-S-H gel);
continues to hydrate, and then begins to remove sulfate from and
some of the ettringite (trisulfate) to form another ■ Through cracks in the paste structure.

86 MARCH 2002 / Concrete international

 If the concrete is moisture-saturated, the pore solution will fill
both the capillary pore system and the cracks in the paste. The
connectivity of these systems is critical in controlling how far Water Capillary
and how fast solutions can move into the concrete. For concrete pores
produced with a w/c of 0.6 and higher, the capillary system will
be continuous (percolating system), and aggressive reactants will
easily move through the concrete mass. With decreasing w/c,
however, the pore system becomes isolated into smaller and
smaller unconnected clusters of pores. These small clusters close
to the surface may become saturated with solution and dissolved Gel
reactants. Any further movement through the paste has to occur
through the C-S-H gel or through the hydration products.

Unhydrated cement
The rate of liquid transport through the C-S-H phase is on the
order of 1000 times slower than that occurring through the
Unhydrated cement
capillary pores. Therefore, even if the capillary pores account for
only 1% of the total transport cross section, they will still provide
90% of the solution transport. Any chemical reactions that
depend solely on transport through C-S-H are negligible for all Fig. 1: Relative volumes of cement, water, and hydration
practical purposes. products before (left) and after (right) hydration
If liquid-filled cracks are present, the movement of a solution
will be proportional to the number and size of cracks. result in expansion, a buildup of internal pressure, and then
Additionally, cracks will also provide shortcuts between pore cracking. Concrete is a brittle material and there-fore can only
clusters and thereby expose deeper layers of concrete to the expand to a limited degree before cracking. In broad terms, the
penetrating solution. With sufficient cracking, the cracks observed expansion is equal to the sum of the crack widths. It is
themselves may form a connected system percolating the not possible to determine the cause of expansion and cracking
concrete and allowing dissolved reactants to move into the from the appearance of the crack pattern on the surface of the
system more quickly. concrete. Interior concrete samples must be examined
The most common type of chemical attack on concrete results microscopically, chemically, or both, to determine the root cause
from exposure to soils or ground water containing elevated of the internal expansion. In field concrete exposed to the
sulfate contents. This malady is a common occurrence in the elements, there are two basic modes of expansion:
western United States, and has been known for many years.
Production of ASTM Type II and Type V cements is particularly ■ The aggregate can expand relative to the cement paste;
intended to provide resistance to deleterious sulfate exposure. and
The mechanism of sulfate attack is relatively simple. If cement ■ The cement paste can expand relative to the aggregate. It
contains an elevated amount of C3A, a substantial amount of the follows from physical considerations that, in a
calcium monosulfoaluminate (AFm) phase will form during composite system consisting of expanding particles in a matrix,
hydration. This substance is reactive, and if additional sulfate cracks are formed in the matrix radiating away from the particles.
from an outside source such as soil or ground water penetrates Expansion of particles in a hardened paste, such as aggregate
the concrete, the monosulfate will readily react with it and particles undergoing ASR, causes the particles to crack and the
convert back to the AFt phase, or ettringite. The conversion of crack to extend outward into the surrounding paste. A particle
the monosulfate phase into ettringite will result in a significant cracking, when expanded from the surface, is actually a fairly
volumetric increase and be disruptive to the concrete. If this is common experience. When ice cubes are dropped into a drink
allowed to continue, the concrete will eventually be destroyed. and a familiar crackling sound is heard—that is the sound of the
ice being heated on the outside, expanding, and then cracking.
The expansion at the surface causes the inner part of the particle
Concrete made at a high w/c and subjected to sodium sulfate- to be under tensile stress, and it cracks from the inside outward.
containing ground water can be damaged by another mechanism.
Studies conducted in California have shown that, even with the
use of ASTM Type II and Type V cement types, if water is able Shrinkage of cement paste is a common phenomenon related
to percolate through the concrete, the wetting and drying on the to hydration, and from a cracking point of view, is equivalent to
surfaces exposed to air result in serious deterioration. Damage is the expansion of aggregate particles.
mainly from the repeated crystallization of alkali sulfates and Consider, however, what would happen if the paste actually
carbonates during the drying cycles (Stark 1989). expands relative to the aggregate particles. When cement paste
expands relative to the aggregate particles, as it has in DEF, gaps
open up around the particles. This concept is actually a little
Chemistry and cracking counterintuitive, in that one might expect the expanding paste to
Cracks in concrete may develop for physical reasons such as actually crush the particle rather than create a gap around it. This
drying shrinkage or mechanical loading. Local chemical is similar to the question of whether a hole that has been drilled
reactions in the concrete, however, may also

Concrete international / MARCH 2002 87

 Paste expansion
ASR is an example of a chemical reaction in which the
aggregate portion of the concrete plays a role in the

Fig. 2: Expansion mind experiment—a) a matrix containing aggregate particles;

b) the whole system is expanded by 20%; and c) the aggregate particles are
returned to their original size, leaving voids around them

(a)
(b) (c)
through a piece of metal will become smaller or larger in
diameter as the metal is heated and expands.
One way to understand the phenomenon is to consider
concrete made with aggregate that does not expand, and conduct
the following imaginary experiment (Fig. 2). First, assume that
the concrete, including both paste and aggregate, expands, say
20%, in all directions. There is no distortion, no cracking, and
the concrete is just somewhat larger. Next, since the aggregate
particles were not expanding, shrink them back to their original
size. What has happened? The particles now rattle around in
holes that are 20% larger than the individual particles. Thus,
gaps have actually formed around the aggregate particles and,
furthermore, these gaps are proportional to particle size.

Alkali-silica reactivity
The chemical reaction called ASR takes place between the
highly alkaline (very high pH) pore solution and reactive
siliceous portions of some aggregate particles. The large amount
of hydroxyl (OH–) ions present in the pore solution, due to a
high alkali concentration (potassium and sodium), dissolve the
reactive silica on the aggregate surfaces to form an alkali silicate
gel. Although any form of silica can react with alkali hydroxides
in theory, it is the siliceous rocks such as opal, greywacke, some
chert, and glassy volcanic materials that appear to be the most
reactive. Reactive siliceous aggregates will form alkali-silica gel
starting at the surface of the aggregate and moving inwards.
Tensile stresses build up during the reaction, causing the
aggregate particles and the surrounding paste to crack. In severe
cases, the cracks will inter-connect and lead to weakening of the
concrete. The weakening is only due to the cracks; the paste
between the cracks maintains its composition and strength.

Denser polycrystalline rocks, such as granites, will react much

more slowly. The chemical reaction will occur at those
heterogeneous areas of grain boundaries. In such cases, only a
minimal degree of reaction may be needed to cause cracking, but
only meager amounts of gel will form. In ASR distress, since
each internal fracture in concrete creates an empty space, the
alkali-silica reaction causes corresponding incremental volume
increases (Helmuth 1993). The resulting visual evidence of the
reaction is the “map cracking” observed on the concrete surface.
 deterioration mechanism. Reactions liquid and subse-quently recrystallize as larger crystals in any
related to sulfate are a group of available spaces, such as cracks and pores. Ostwald ripening is
reactions that involve only the cement
paste. DEF reactions that are associated a general chemical principle, and calcium hydroxide
with concrete exposure to high crystallization behaves in a similar fashion.
temperatures during curing, in systems Regarding DEF, however, a high concrete temperature at an
containing normal-sulfate-content early age is a very important parameter. At certain temperatures,
cements, have been a “hot button” topic generally above 70 °C, and more frequently above 80 °C,
of discussion for some time. DEF results ettringite becomes unstable because its solubility increases. This
from a chemical reaction, or sets of temperature is strongly depen-dent upon the alkali content and
reactions, that are still not well other compositional factors of the cement that are less well
understood at the present time, but the understood. Where the components of the ettringite go after its
diagnosis is relatively decomposition is not clear. Portions of ettringite may be
consumed by the C-S-H or may stay in solution. This issue is the
simple; deterioration results from paste expansion in the subject of much scientific research and discussion at present
affected concrete. (Tennis et al. 1999). What is clear, however, is that the
In DEF-affected concrete, the AFt phase, or ettringite, is chemistry matters, even if we do not fully understand the precise
usually observed. But ettringite in concrete is not unique to sequence of reactions.
DEF. Ettringite is a normal hydration product formed by the
chemical reaction between the aluminate phases of cement, One sign of paste expansion is the presence of voids or
water, and calcium sulfate (gypsum), as mentioned previously. cracks around the aggregate particles, as explained previously.
The formation of ettringite takes place in the paste and is Usually, ettringite fills these gaps. There are also examples of
uniformly distributed. Within mature concrete exposed to moist paste expansion showing empty gaps, however.
conditions, ettringite is usually found in pores and cracks. This The following is one possible explanation of the observed
is not an indication of damage, but rather the result of paste expansion related to DEF, in which sulfate dissolved in
the pore liquid reacts with the anhydrous and hydrated
a normal recrystallization process known as “Ostwald ripening.” aluminate particles in the hardened cement paste. Johansen
This means that small crystals have a higher solubility than large and Thaulow (1999) discussed this previously. Hardened
crystals, and when concrete become water-saturated to a certain paste, mortar, and heat-treated as well as normally cured
concrete are all places where one can find unhydrated clinker
degree, the small crystals within the paste dissolve in the pore
particles

88 MARCH 2002 / Concrete international

in various amounts. As mentioned previously, above certain Ettringite,” Ettringite: The Sometimes Host of Destruction, SP-177, B.
temperatures ettringite is unstable, and the primary hydrated Erlin, ed., American Concrete Institute, Farmington Hills, Mich., 1999, pp.
aluminate phase is the calcium monosulfoaluminate (AFm). 47-64.
Therefore, after cooling to room temperature following heat Kosmatka, S. H. and Panarese, W. C., “Design and Control of
Concrete Mixtures,” PCA Engineering Bulletin EB001.13T, Portland
treatment, concrete will contain anhydrous aluminate particles,
Cement Association, Skokie, Ill., 1994.
anhydrous aluminate particles with AFm, and AFm phases. Le Chatelier, H., Experimental Researches on the Constitution of
During the passage of time and moist curing of such concrete, Hydraulic Mortars, Translated by J. L. Mack, McGraw Publishing Co., New
these particles will continue to react with sulfate in the pore York, N.Y., 1905.
solution. Both ettringite and AFm phases form, depending Stark, D., “Durability of Concrete in Sulfate-Rich Soils,” PCA
upon the composition of the pore solution. The sulfate, Research and Development Bulletin RD097, Portland Cement
liberated from the C-S-H that had initially absorbed it during Association, Skokie, Ill., 1989.
the heat treatment, maintains the sulfate concentration of the Steinour, H. H., “The Setting of Portland Cement: A Review of
pore solution. Theory, Performance and Control,” PCA Research Department Bulletin
98, Portland Cement Association, Chicago, Ill., Nov. 1958.
Tennis, P. D.; Bhattacharja, S.; Klemm, W. A.; and Miller, F. M.,
The hardened paste confines the reacting particles and the “Assessing the Distribution of Sulfate in Portland Cement and Clinker and
volume of the AFm and ettringite (AFt) formed will result in Its Influence on Expansion in Mortar,” Cement, Concrete, and Aggregates,
development of localized pressure. Crystals under pressure have V. 21, No. 2, Dec. 1999, pp. 212-216.
a higher solubility than what they would have when not under Weaver, W. S., “Committee C-1 on Cement—Seventy-Five Years of
pressure. When more and more AFm and AFt phases are formed Achievement,” Cement Standards—Evolution and Trends, ASTM STP 663,
on the reacting particles, the pressure will increase and, P. K. Mehta, ed., American Society for Testing and Materials, Philadelphia,
therefore, so will their solubility. If the solubility increases to a Pa., 1978, pp. 3-15.
level corresponding with the actual concentration in the pore
solution, the crystal growth stops and the pressure will act Received and reviewed under Institute publication policies.
locally on the particle and its surroundings. In this way, the
reacting particle can act as a local pressure center. This will
cause stress to build up in the surrounding paste as sort of a
“sphere of influence” around the particle. If the pressure created
is larger than the tensile strength of the paste, the paste will
crack or yield. If the reacting particles are sufficiently close to
each other, mass volu-metric expansion will result.
ACI member Vagn C. Johansen is a Senior
Principal at Construction Technology
Laboratories (CTL), Inc. He has more than
20 years of experience with cement
Conclusion
chemistry and cement manufacture, and 10
Returning to the original question, why does chemistry years of experience with concrete materials
matter? The answer can be summarized as follows: chemistry and durability-related work. Previously, he
matters because concrete composition and performance are worked in Denmark for Idorn Consult, and
based upon a variety of chemical reactions that range from the with F.L. Smith & Co.
original setting and hardening of the portland cement
constituent to the eventual desired engineering properties. The
durability of concrete depends on chemical processes Waldemar A. Klemm is an Affiliated
developing out of cement and aggregate compositional factors, Consultant with CTL, Inc. He has 30 years of
curing conditions, and exposure to a variety of envi-ronmental experience with the U.S. cement industry in
effects. The chemical reactions that occur during the hydration plant process, chemistry, research, and
of the clinker minerals determine the concrete microstructure. developmental activities. He has authored
The hardened concrete is chemically reactive given the right numerous technical reports and scientific
conditions, as shown in the examples. Therefore, it is essential papers on clinkering chemistry, cement
to design concrete mixtures properly and erect structures in a hydration, admixture research, cement
way to control or adequately compensate for chemical manufacturing, and environmental problems.
reactivity. He is a member of ASTM.

ACI member Peter C. Taylor has been at

CTL, Inc., since 1997, and is a Senior
References Engineer. He is a graduate in Civil
Bogue, R. H. and Steinour, H. H., “Origin of the Special Chemical Engineering from the University of Cape
Symbols Used by Cement Chemists,” Journal of the PCA Research & Town, South Africa, and has 18 years’
Development Laboratories, V. 3, No. 3, Sept. 1961, pp. 20-21. experience in consulting and research. He
Helmuth, R., Alkali-Silica Reactivity: An Overview of Research, is a member of ACI Committees 232, Fly
SHRP-C-342, Strategic Highway Research Program, National Research Ash and Natural Pozzolans in Concrete,
Council, Washington, D.C., 1993. and 236, Material Science of Concrete.
Johansen, V. and Thaulow, N., “Heat Curing and Late Formation of

Concrete international / MARCH 2002 89