FORUM
The Use of Chemical Admixtures in Concrete.
Part I: Admixture-Cement Compatibility
Claude Bedard, P.E.
Vice President and General Manager, North American Admixtures
Division, Euclid Chemical Co., 19218 Redwood Rd., Cleveland, OH
Downloaded from ascelibrary.org by New York University on 05/18/15. Copyright ASCE. For personal use only; all rights reserved.
44110-2799. E-mail: c.bedard@qc.aira.com
Noel P. Mailvaganam
Principal Scientist, National Research Council of Canada NRC,
Montreal Road, Bldg. M-20, Ottawa, Ontario K1A 0R6, Canada. E-mail:
noel.mailvaganam@nrc.ca
Introduction
Todays concrete incorporates a mixture of chemical and mineral
admixtures. They interact with cement constituents and influence
cement hydration. An admixtures performance is dependent on
its type and dosage, composition, specific surface area of the ce-
ment, type and proportions of the aggregate, sequence of addition
of water and admixture, compatibility of admixtures, water/
cement ratio, and temperature and conditions of curing. The use
of a mixture of admixtures in a mix raises further compatibility
issues, which need rectifying before field use. Incompatibilities
between admixtures will be discussed in a future forum contribu-
tion, along with the resolution of related problems.
Admixture-Cement Interactions
The progress of the hydration reaction with time involves five
stages see Fig. 1:
Iinitial hydration processes 015 min;
IIinduction period or lag phase 15 min4 h;
IIIacceleration and setting 48 h;
IVdeceleration and hardening 824 h; and
Vcuring 128 days.
In Stage I, water wets the cement particles and solubilizes the
cement phases. After initial solubilization, the formation of the
solid hydration products hydrates occurs from the solution
phase or at a solid-solution interface. Hydration products grow at
a rate determined by the water and by ionic species availability at
the reaction sites. The presence of admixtures that can interfere
with the growth processes influences the hydration reaction rate,
the reaction products, or both. In the latter part of Stage I, the
cement particles in the paste become coated with a layer of hy-
drates that sharply reduces the rate of the reactions. The system
enters a dormant period Stage II; processes initiated during
Stage I, however, continue throughout the dormant period.
In the early part of Stage II, reactions of the aluminate and
gypsum phases play a predominant role in the setting of the paste.
If the solubilization of gypsum to produce sulfate ions is too
low, flash set may occur. Flash set is distinguished from false set
in that it evolves considerable heat and the rigidity of the mix
JOURNAL OF PERFORMANCE OF CONSTRUCTED FACILITIES ASCE / NOVEMBER 2005 / 263
J. Perform. Constr. Facil. 2005.19:263-266.
cannot be dispelled by further mixing without adding water.
False set is caused by the presence of hemihydrate or anhydrite. If
the solubilization of gypsum is too high because of the presence
of the hemihydrate form of gypsum & sodium and potassium
sulfates, then extensive growth of gypsum crystals occurs, result-
ing in false set. The presence of an adequate sulfate ion content,
however, increases the production of the cementing gel. Since
these processes determine the workability and setting characteris-
tics of the concrete, the interaction of a chemical admixture with
any of the reactive species aluminate and gypsum, or its inter-
ference with diffusion, nucleation, and growth processes, can sig-
nificantly influence the behavior of concretes during the dormant
period. Near the end of the dormant period, the rate of cement
hydration increases sharply because of the breakdown of the hy-
drate layer, nucleation, and the growth of hydrates and calcium
hydroxide, leading to setting and hardening.
Admixture-Cement Incompatibility:
Selected Examples
Chemical admixtures affect the properties of concrete in diverse
ways, and more than one effect can occur at a time e.g., chemical
interference with hydration reactions or physical interaction with
the hydration products. These effects result in the alteration of
the rate of hydration Fig. 2 or in the composition and morphol-
ogy of the hydrated products Fig. 3. Admixtures that modify the
properties of fresh concrete may cause problems through early
stiffening or undesirable retardation of the setting time. Early
stiffening is often caused by changes in the rate of reaction be-
tween tricalcium aluminate and sulfate in the cement. Retardation
can result from an overdose of admixture or from a decrease in
the ambient temperature. Both delay calcium silicates hydration.
As more chemicals are added to the mix, compatibility with
the cement and other admixtures becomes the central parameter
governing the selection. Side effects or reactions among chemi-
cals attributable to the sequence of addition, cement type,
temperature change, and batching equipment can all affect
performance.
Fig. 1. Stages in hydration reaction Jolicoeur et al. 1994, with
permission

Fig. 2. Alteration in rate of hydration at three dosages of the
Downloaded from ascelibrary.org by New York University on 05/18/15. Copyright ASCE. For personal use only; all rights reserved.
chemical admixture Jolicoeur et al. 1994, with permission
Abnormal Setting and Loss of Workability Attributable
to Type of Gypsum Present in Cement
The different mineral phases of cement react with water at differ-
ent rates. Calcium aluminate phases react faster than silicates.
Because of the high reactivity of calcium aluminate, its hydration
reaction occurs in the presence of sulfate ions provided by adding
gypsum calcium sulfate to the clinker. Gypsum plays a crucial
role in cement hydration, particularly setting, and the influence of
chemical admixtures on any process where it is involved may be
significant.
Most of the calcium sulfate introduced into cement is present
in one or more forms, gypsum CaSO4 2 H2O, hemihydrate
CaSO4 1 / 2 H2O, soluble anhydrite, or natural anhydrite
CaSO4. It is important that the calcium sulfate be sufficiently
soluble in the cement/water aqueous phase to provide calcium and
sulfate ions for the formation of the initial hydrates and to prevent
flash set. Dehydration of gypsum when it is interground with too
hot a clinker can produce these two forms of gypsum. When such
cement is mixed with water, the hemihydrate and anhydrite hy-
drate to gypsum. A rapid set takes place, resulting in stiffening of
the paste.
Gypsum is either partially replaced or often substituted by
natural anhydrite in the manufacture of portland cement. Portland
cement containing natural anhydrite performs normally in the ab-
sence of admixtures. However, using chemical admixtures with
such cements produces problems. The natural anhydrite, whether
Fig. 3. Alteration in morphology of hydrates: a cement particle
agglomerates; and b dispersion of cement particles by admixture
Rixon and Mailvaganam 1995, with permission
264 / JOURNAL OF PERFORMANCE OF CONSTRUCTED FACILITIES ASCE / NOVEMBER 2005
J. Perform. Constr. Facil. 2005.19:263-266.
Fig. 4. Comparative effect of calcium lignosulfonate admixture
CIS on gypsum and natural anhydrite Dodson 1990
interground purposely or as an impurity in gypsum, can produce
rapid set in paste, mortar, or concrete in the presence of calcium
lignosulfonate CLS admixtures.
The rate of solution of natural anhydrite is slower than that of
gypsum or calcium sulfate hemihydrate and is further retarded in
the presence of chemical admixtures, often producing rapid set
and an increase in the rate of concrete slump loss Fig. 4. Chemi-
cal admixtures that are based on salts of hydroxylated carboxylic
acids as well as carbohydrates sugars, corn syrups act like CLS
when added to concrete containing natural anhydrite.
Influence of Soluble Alkalies on Stability of Air Void
Structure
Previous investigations indicated that increased soluble alkalis
sodium and potassium oxides can produce alkali-induced break-
down of the air void structure AVS, making concrete more vul-
nerable to freeze-thaw damage. Recent work, however, shows
that soluble alkalis do not influence the production of the AVS.
Only when high sulfate levels exist inherent or by extraneous
addition is AVS production difficult. An increase in the soluble
alkali content of cement can significantly improve the stability of
AVS, particularly when a superplasticizer SP is added to the
concrete.
Suppressed Strength Development in Sulfate-Resistant
Concrete
Concrete made with Type 50 cement, which has low calcium
aluminate content, uses low dosages of CLS-based water reducing
agents WRAs and SPs. Problems arise when workers trying to
cope with placing and pumping difficulties add further WRA and
SP without knowledge of the synergistic effect produced on the
concretes plastic properties by admixture combinations, particu-
larly with Type 50 cement.
In some instances, the addition of higher than normal WRA
dosages or a combination of WRA and SP in air-entrained con-
crete containing Type 50 cement with unusually low calcium alu-
minate contents 1 % , can lead to severe retardation of strength
development and, in drastic cases, the arrest of hydration.
Cement-Superplasticizer Incompatibility
in High-Performance Concrete
Some cement-SP combinations cause such problems as low flu-
idification effect, rapid slump loss, severe segregation, loss of
entrained air, and extended set retardation. These are briefly re-
viewed below.
 Loss of Workability in High-Performance Concrete
Cement-SP incompatibility problems are magnified in high-
performance concrete HPC because of the much-reduced water/
cement ratios and concomitantly higher SP doses. Cement-SP
compatibility is critical in making HPC, such that some cements
are rejected not because of the difficulty in achieving the required
strength but rather because of the very rapid slump loss.
The principal cement and admixture factors influencing rheo-
logical behavior in HPC-SP systems are in the cements trical-
cium aluminate C3A content, Blaine fineness, solubility of the
gypsum, and composition of the SP. In addition to the physical
role of cement particle dispersers, the admixture composition af-
fects the rate of cement hydration and the solubility of gypsum.
The calcium sulfate content of portland cement is regulated for
Downloaded from ascelibrary.org by New York University on 05/18/15. Copyright ASCE. For personal use only; all rights reserved.

compliance with national standards, primarily on the basis of the

workability and strength behavior of standard pastes and mortars
made at water/cement W / C ratios of approximately 0.5. It pro-
vides satisfactory results for commonly used concretes at W / C
ratios greater than 0.4 with compressive strengths in the range of
20 to 40 MPa. HPCs, however, have lower W / C 0.2 to 0.4, with
compressive strengths ranging from 50 to 100 MPa. In such con-
cretes, the rate of solubility of gypsum becomes a limiting factor
in controlling workability, since there is less water to solubilize
the gypsum.

Segregation of Mix
Segregation of the mix can occur when larger dosages of the
admixture are added to offset the lack of initial fluidizing of the
concrete or to reinstate workability. The saturation point for the
admixture varies for different cement types and cement blends
for example, Type 50 cement and blends of fly ash and portland
cement require much less admixture to produce the fluidizing
effect than that required for a Type 10 cement because of the
smaller amount of tricalcium aluminate C3A contained in Type
50 cement compared with Type 10 and the reduction of the C3A
content in blended mixtures. A mix showing signs of segregation
can be recovered by adding extra quantities of cement to the mix
to consume the excess of SP.
Destabilization of Air System in Superplasticized Concrete
The addition of an SP modifies the relationship between the air
content and the spacing factor air-bubble spacing and sometimes
significantly destabilizes the air-void system AVS Fig. 5. Low
dosages of SPs appear to be less problematic than higher doses.
The possible cause for this decrease is the escape of air from the
fluidized concrete and the coalescence of the air bubbles.
For ready-mix concrete, the instability of the air-void system
must be linked to a practical measurement of the air content. The
only way to assess the stability of the AVS in superplasticized
concrete is to carry out a preliminary test for each cement-AEA
air entraining agent-SP combination. Several tests of a specific
combination of cement-AEA-SP may yield a relationship between
the air content and spacing factor. Any change of AEA type or
brand of cement will necessitate new evaluations.
Abnormal Retardation of Superplasticized Mixes
Extended retardation of superplasticized mixes is closely related
to Blaine surface area, chemical composition mainly the C3A
content of the cement, SP dosage, and prevailing ambient and
concrete temperatures. Low C3A cements, such as Type 50 ce-
ments, are susceptible to extension of their set, even at dosages
that are considered normal for other cements with higher C3A
contents.
Fig. 5. Effect of mixing time after addition of air-entraining agent
Mix 3.1 in combination with superplasticizer Mix 3.2 and in com-
bination with water reducer and superplasticizer Mix 3.3 on total air
content: a in fresh concrete; and b in hardened concrete Mark and
Gjoerv 1997
The following can identify whether the compatibility problem
is attributable to the reactivity of the cement or the poor perfor-
mance of the admixture:
Check the reactivity of the cement C3A content, type of gyp-
sum, and Blaine fineness.
Determine performance of the cement-SP combinations by
cross-testing with other SPs and cements.
Try small doses of a hydroxycarboxylic-based retarder.
Try blending the cement with fly ash or slag.
If these measures are unsuccessful, then the W / C should be
progressively increased while varying the dosage of the SP to
achieve the desired workability and required strength.
Summary
It is important to understand admixture-cement and admixture-
admixture interactions so optimal use of materials can be
achieved, admixture-cement incompatibility can be prevented,
better troubleshooting of field problems is possible, and more
accurate prediction of concrete properties is possible.
Bibliography
Baalbaki, M., and Aitcin, P. C. 1994. Cement superplasticizer/
air entraining agent compatibility. 4th Int. CANMET/ACI
Conf. on Superplasticizers and Other Chemical Admixtures,
Montreal, Que., Canada, 47-62.

JOURNAL OF PERFORMANCE OF CONSTRUCTED FACILITIES ASCE / NOVEMBER 2005 / 265

J. Perform. Constr. Facil. 2005.19:263-266.

 Dodson, V. 1990. Concrete admixtures, Van Nostrand Reinhold,
New York, 34-38.
Johnston, C. D. 1987. Admixture-cement incompatibilityA
case history. Concr. Int., 99, 4351.
Jolicoeur, C., Nkinamubanzi, P. C., Simard, M. A., and Picotte,
M. 1994. Progress in understanding the functional proper-
ties of superplasticizers in fresh concrete, 4th CANMET/ACI
Int. Conf. on Superplasticizers and Other Chemical Admix-
tures in Concrete, ACI SP-148, V. M. Malhotra, ed., American
Concrete Institute, Farmington Hills, Mich., 6388.
Mailvaganam, N. P. 1998. Admixture compatibility in special
concrete. Int. Conf. on High Performance Concrete, Gra-
mado, Brazil, pp. 121-133.
Downloaded from ascelibrary.org by New York University on 05/18/15. Copyright ASCE. For personal use only; all rights reserved.
Manabe, T., and Kawada, N. 1960. Abnormal setting of cement
paste owing to calcium lignosulfonate. Semento Konkurito,
162, 2427.
Mark, J. H., and Gjoerv, O. E. 1997. Effect of gypsum-
hemihydrate ratio in cement on rheological properties of fresh
concrete. ACI Mater. J., MarchApril, 142146.
Plante, P., and Pigeon, M. 1983. Air void stability in the pres-
ence of superplasticizers as affected by soluble alkalis. Proc.,
3rd Int. Symp. on Superplasticizers and Other Chemical Ad-
mixtures in Concrete, ACI SP-119, Detroit, 118143.
Rixom, M. R. and Mailvaganam, N. P. 1995. Early hydration
reactions in portland cement concrete. Holderchem Workshop
on Admixture Applications, Zurich, Switzerland, 2337.

266 / JOURNAL OF PERFORMANCE OF CONSTRUCTED FACILITIES ASCE / NOVEMBER 2005

J. Perform. Constr. Facil. 2005.19:263-266.