1. What is cement?

The importance of concrete in modern society cannot be underestimated.

Cement, as it is commonly known, is a mixture of compounds made by burning limestone and clay
together at very high temperatures ranging from 1400 to 1600C.
Although there are other cements for special purposes, this module will focus solely (only) on portland
cement and its properties. The production of portland cement begins with the quarrying of limestone,
CaCO3. Huge crushers break the blasted limestone into small pieces. The crushed limestone is then mixed
with clay (or shale), sand, and iron ore and ground together to form a homogeneous powder. However, this
powder is microscopically heterogeneous. (See flowchart.)
Water is the key ingredient (thanh phan)
Cement paste will binds (dnh) the aggregates together. The water causes the hardening of concrete through
a process called hydration
Aggregates should be clean, hard, and strong. The aggregate is usually washed to remove any dust, silt,
clay, organic matter, or other impurities that would interfere with the bonding reaction with the cement
paste
tricalcium aluminate and tetracalcium aluminoferrite also react with water, and reaction between them
and gypsum is more complicated
The strength of concrete is very much dependent upon the hydration reaction just discussed

Main Constituents in a Typical Portland Cement (Mindess and Young, 1981)

Chemical Name

Chemical Formula

Shorthand
Notation

Percent by
Weight

Tricalcium Silicate

3CaO.SiO2

C3S

50

Dicalcium Silicate

2CaO.SiO2

C 2S

25

Tricalcium Aluminate

3CaO.Al2O3

C3A

12

Tetracalcium Aluminoferrite

4CaO.Al2O3.Fe2O3

C4AF

Gypsum (Thach cao)

CaSO4.H2O

CSH2

3.5

Portland cement consists of five major compounds and a few minor compounds. The composition of
a typical portland cement is listed by weight percentage.

Cement Compound

Weight Percentage Chemical Formula

Tricalcium silicate

50 %

Ca3SiO5 or 3CaO.SiO2

Dicalcium silicate

25 %

Ca2SiO4 or 2CaO.SiO2

Tricalcium aluminate

10 %

Ca3Al2O6 or 3CaO .Al2O3

Tetracalcium aluminoferrite

10 %

Ca4Al2Fe2O10 or 4CaO.Al2O3.Fe2O3

Gypsum

5%

CaSO4.2H2O

Tricalcium silicate (C3S). Hydrates and hardens rapidly and is largely responsible for initial
set and early strength. Portland cements with higher percentages of C3S will exhibit higher
early strength.

Dicalcium silicate (C2S). Hydrates and hardens slowly and is largely responsible for strength
increases beyond one week.

Tricalcium aluminate (C3A). Hydrates and hardens the quickest. Liberates (releases, emit) a
large amount of heat almost immediately and contributes somewhat to early trength. Gypsum
is added to portland cement to retard (lam tre) C3A hydration. Without gypsum, C3A
hydration would cause portland cement to set almost immediately after adding water.
Tetracalcium aluminoferrite (C4AF). Hydrates rapidly but contributes very little to strength.
Its use allows lower kiln temperatures in portland cement manufacturing. Most portland
cement color effects are due to C4AF.

Accelerators: Admixtures that decrease the setting time by increasing the rate of hydration.
Admixture: A material other than water, aggregates, or cement that is used as an ingredient of concrete or
mortar to control setting and early hardening, workability, or to provide additional cementing properties.
Aggregate: Inert solid bodies such as crushed rock, sand, gravel.
Binder: Hardened cement paste.
Bleed: To have water seep to the surface of the cement paste due to settling.
Calcination: Decomposition due to the loss of bound water and carbon dioxide.
Cement: Finely powdered mixtures of inorganic compounds which when combined with water hardens with
hydration.
Cement paste: Cement plus water. When the mass has reacted with water and developed strength it is
called hardened cement paste.
Clay: Type of soil consisting of very fine particles.
Clinker: The material that emerges from the cement kiln after burning. It is in the form of dark, porous
nodules which are ground with a small amount of gypsum to give cement.
Compression: Forces acting inwardly on a body.
Concrete: A hard compact building material formed when a mixture of cement, sand, gravel, and water
undergoes hydration.
Cure: To keep concrete moist during initial hardening.
Deformation: The process of changing the dimensions of a structure by applying a force.
Dormancy period: Time period that concrete retains it workability.
Elasticity: The ability of a material to return to its original shape after being stretched.
Forms: Holders in which concrete is placed to harden.

Gypsum: Calcium sulfate dihydrate, CaSO4.2H2O added to cement to regulate setting.

Hydration: The reaction of cement with water to form a chemical compound.
Kiln: High temperature oven.
Limestone: Mineral rock of calcium carbonate.
Mortar: Cement paste mixed with sand.
Pozzolan cement: Volcanic rock powdered and used in making hydraulic cement.
Porosity: The amount of empty space in concrete.
Portland cement: A cement consisting predominantly of calcium silicates which reacts with water to form a
hard mass.
Retardants: Admixtures that increase the setting time by slowing down hydration.
Set: Transformation of cement paste or concrete from a fluid-like consistency to a stiff mass.
Slump test: Test used to determine workability.
Tension: The stress resulting from elongation.
Workability: How easily fresh concrete can be placed and consolidated in forms

Scientific Principles
What is in This Stuff?
The importance of concrete in modern society cannot be underestimated. Look around
you and you will find concrete structures everywhere such as buildings, roads, bridges,
and dams. There is no escaping the impact concrete makes on your everyday life. So what
is it?

Concrete is a composite material which is made up of a filler and a binder. The binder
(cement paste) "glues" the filler together to form a synthetic conglomerate. The
constituents used for the binder are cement and water, while the filler can be fine or
coarse aggregate. The role of these constituents will be discussed in this section.
Cement, as it is commonly known, is a mixture of compounds made by burning
limestone and clay together at very high temperatures ranging from 1400 to 1600
[[ring]]C.
Although there are other cements for special purposes, this module will focus solely on
portland cement and its properties. The production of portland cement begins with the
quarrying of limestone, CaCO3. Huge crushers break the blasted limestone into small
pieces. The crushed limestone is then mixed with clay (or shale), sand, and iron ore and
ground together to form a homogeneous powder. However, this powder is
microscopically heterogeneous. (See flowchart.)

Figure 1: A flow diagram of Portland Cement production.

The mixture is heated in kilns that are long rotating steel cylinders on an incline. The
kilns may be up to 6 meters in diameter and 180 meters in length. The mixture of raw
materials enters at the high end of the cylinder and slowly moves along the length of the
kiln due to the constant rotation and inclination. At the low end of the kiln, a fuel is
injected and burned, thus providing the heat necessary to make the materials react. It can
take up to 2 hours for the mixture to pass through the kiln, depending upon the length of
the cylinder.

Figure 2: Schematic diagram of rotary kiln.

As the mixture moves down the cylinder, it progresses through four stages of
transformation. Initially, any free water in the powder is lost by evaporation. Next,
decomposition occurs from the loss of bound water and carbon dioxide. This is called
calcination. The third stage is called clinkering. During this stage, the calcium silicates
are formed. The final stage is the cooling stage.
The marble-sized pieces produced by the kiln are referred to as clinker. Clinker is
actually a mixture of four compounds which will be discussed later. The clinker is cooled,
ground, and mixed with a small amount of gypsum (which regulates setting) to produce
the general-purpose portland cement.
Water is the key ingredient, which when mixed with cement, forms a paste that binds the
aggregate together. The water causes the hardening of concrete through a process called
hydration. Hydration is a chemical reaction in which the major compounds in cement
form chemical bonds with water molecules and become hydrates or hydration products.
Details of the hydration process are explored in the next section. The water needs to be
pure in order to prevent side reactions from occurring which may weaken the concrete or
otherwise interfere with the hydration process. The role of water is important because the
water to cement ratio is the most critical factor in the production of "perfect" concrete.
Too much water reduces concrete strength, while too little will make the concrete
unworkable. Concrete needs to be workable so that it may be consolidated and shaped
into different forms (i.e.. walls, domes, etc.). Because concrete must be both strong and
workable, a careful balance of the cement to water ratio is required when making
concrete.

Aggregates are chemically inert, solid bodies held together by the cement. Aggregates
come in various shapes, sizes, and materials ranging from fine particles of sand to large,
coarse rocks. Because cement is the most expensive ingredient in making concrete, it is
desirable to minimize the amount of cement used. 70 to 80% of the volume of concrete is
aggregate keeping the cost of the concrete low. The selection of an aggregate is
determined, in part, by the desired characteristics of the concrete. For example, the
density of concrete is determined by the density of the aggregate. Soft, porous aggregates
can result in weak concrete with low wear resistance, while using hard aggregates can
make strong concrete with a high resistance to abrasion.
Aggregates should be clean, hard, and strong. The aggregate is usually washed to remove
any dust, silt, clay, organic matter, or other impurities that would interfere with the
bonding reaction with the cement paste. It is then separated into various sizes by passing
the material through a series of screens with different size openings.
Refer to Demonstration 1
Table 1: Classes of Aggregates
class

examples of
aggregates used

uses

ultralightweight

vermiculite
ceramic spheres
perlite

lightweight concrete which can be sawed or

nailed, also for its insulating properties

lightweight

expanded clay
shale or slate
crushed brick

used primarily for making lightweight concrete

for structures, also used for its insulating
properties.

crushed limestone
sand
normal weight river gravel
used for normal concrete projects
crushed recycled
concrete
heavyweight

steel or iron shot used for making high density concrete for
steel or iron pellets shielding against nuclear radiation

Refer to Demonstration 2
The choice of aggregate is determined by the proposed use of the concrete. Normally
sand, gravel, and crushed stone are used as aggregates to make concrete. The aggregate
should be well-graded to improve packing efficiency and minimize the amount of cement
paste needed. Also, this makes the concrete more workable.
Refer to Demonstration 3
Properties of Concrete

Concrete has many properties that make it a popular construction material. The correct
proportion of ingredients, placement, and curing are needed in order for these properties
to be optimal.
Good-quality concrete has many advantages that add to its popularity. First, it is
economical when ingredients are readily available. Concrete's long life and relatively low
maintenance requirements increase its economic benefits. Concrete is not as likely to rot,
corrode, or decay as other building materials. Concrete has the ability to be molded or
cast into almost any desired shape. Building of the molds and casting can occur on the
work-site which reduces costs.
Concrete is a non-combustible material which makes it fire-safe and able withstand high
temperatures. It is resistant to wind, water, rodents, and insects. Hence, concrete is often
used for storm shelters.
Concrete does have some limitations despite its numerous advantages. Concrete has a
relatively low tensile strength (compared to other building materials), low ductility, low
strength-to-weight ratio, and is susceptible to cracking. Concrete remains the material of
choice for many applications regardless of these limitations.
Hydration of Portland Cement
Concrete is prepared by mixing cement, water, and aggregate together to make a
workable paste. It is molded or placed as desired, consolidated, and then left to harden.
Concrete does not need to dry out in order to harden as commonly thought.
The concrete (or specifically, the cement in it) needs moisture to hydrate and cure
(harden). When concrete dries, it actually stops getting stronger. Concrete with too little
water may be dry but is not fully reacted. The properties of such a concrete would be less
than that of a wet concrete. The reaction of water with the cement in concrete is
extremely important to its properties and reactions may continue for many years. This
very important reaction will be discussed in detail in this section.
Portland cement consists of five major compounds and a few minor compounds. The
composition of a typical portland cement is listed by weight percentage in Table 2.
Cement Compound

Weight Percentage Chemical Formula

Tricalcium silicate

50 %

Ca3SiO5 or 3CaO.SiO2

Dicalcium silicate

25 %

Ca2SiO4 or 2CaO.SiO2

Tricalcium aluminate

10 %

Ca3Al2O6 or 3CaO .Al2O3

Tetracalcium aluminoferrite

10 %

Ca4Al2Fe2O10 or 4CaO.Al2O3.Fe2O3

Gypsum

5%

CaSO4.2H2O

Table 2: Composition of portland cement with chemical composition and weight percent.

When water is added to cement, each of the compounds undergoes hydration and
contributes to the final concrete product. Only the calcium silicates contribute to strength.
Tricalcium silicate is responsible for most of the early strength (first 7 days). Dicalcium
silicate, which reacts more slowly, contributes only to the strength at later times.
Tricalcium silicate will be discussed in the greatest detail.
The equation for the hydration of tricalcium silicate is given by:
Tricalcium silicate + Water--->Calcium silicate hydrate+Calcium hydroxide + heat
2 Ca3SiO5 + 7 H2O ---> 3 CaO.2SiO2.4H2O + 3 Ca(OH)2 + 173.6kJ
Upon the addition of water, tricalcium silicate rapidly reacts to release calcium ions,
hydroxide ions, and a large amount of heat. The pH quickly rises to over 12 because of
the release of alkaline hydroxide (OH-) ions. This initial hydrolysis slows down quickly
after it starts resulting in a decrease in heat evolved.
The reaction slowly continues producing calcium and hydroxide ions until the system
becomes saturated. Once this occurs, the calcium hydroxide starts to crystallize.
Simultaneously, calcium silicate hydrate begins to form. Ions precipitate out of solution
accelerating the reaction of tricalcium silicate to calcium and hydroxide ions. (Le
Chatlier's principle). The evolution of heat is then dramatically increased.
The formation of the calcium hydroxide and calcium silicate hydrate (C-S-H) crystals
provide "seeds" upon which more calcium silicate hydrate can form. The calcium silicate
hydrate crystals grow thicker making it more difficult for water molecules to reach the
unhydrated tricalcium silicate. The speed of the reaction is now controlled by the rate at
which water molecules diffuse through the calcium silicate hydrate coating. This coating
thickens over time causing the production of calcium silicate hydrate to become slower
and slower.

Figure 3: Schematic illustration of the pores in calcium silicate through different stages
of hydration.

The above diagrams represent the formation of pores as calcium silicate hydrate is
formed. Note in diagram (a) that hydration has not yet occurred and the pores (empty
spaces between grains) are filled with water. Diagram (b) represents the beginning of
hydration. In diagram (c), the hydration continues. Although empty spaces still exist, they
are filled with water and calcium hydroxide. Diagram (d) shows nearly hardened cement
paste. Note that the majority of space is filled with calcium silicate hydrate. That which is
not filled with the hardened hydrate is primarily calcium hydroxide solution. The
hydration will continue as long as water is present and there are still unhydrated
compounds in the cement paste.
Dicalcium silicate also affects the strength of concrete through its hydration. Dicalcium
silicate reacts with water in a similar manner compared to tricalcium silicate, but much
more slowly. The heat released is less than that by the hydration of tricalcium silicate
because the dicalcium silicate is much less reactive. The products from the hydration of
dicalcium silicate are the same as those for tricalcium silicate:
Dicalcium silicate + Water--->Calcium silicate hydrate + Calcium hydroxide +heat
2 Ca2SiO4 + 5 H2O---> 3 CaO.2SiO2.4H2O + Ca(OH)2 + 58.6 kJ
The other major components of portland cement, tricalcium aluminate and tetracalcium
aluminoferrite also react with water. Their hydration chemistry is more complicated as
they involve reactions with the gypsum as well. Because these reactions do not contribute
significantly to strength, they will be neglected in this discussion. Although we have
treated the hydration of each cement compound independently, this is not completely
accurate. The rate of hydration of a compound may be affected by varying the
concentration of another. In general, the rates of hydration during the first few days
ranked from fastest to slowest are:
tricalcium aluminate > tricalcium silicate > tetracalcium aluminoferrite > dicalcium
silicate.
Refer to Demonstration 4
Heat is evolved with cement hydration. This is due to the breaking and making of
chemical bonds during hydration. The heat generated is shown below as a function of
time.

Figure 4: Rate of heat evolution during the hydration of portland cement

The stage I hydrolysis of the cement compounds occurs rapidly with a temperature
increase of several degrees. Stage II is known as the dormancy period. The evolution of
heat slows dramatically in this stage. The dormancy period can last from one to three
hours. During this period, the concrete is in a plastic state which allows the concrete to be
transported and placed without any major difficulty. This is particularly important for the
construction trade who must transport concrete to the job site. It is at the end of this stage
that initial setting begins. In stages III and IV, the concrete starts to harden and the heat
evolution increases due primarily to the hydration of tricalcium silicate. Stage V is
reached after 36 hours. The slow formation of hydrate products occurs and continues as
long as water and unhydrated silicates are present.
Refer to Demonstration 5
Strength of Concrete
The strength of concrete is very much dependent upon the hydration reaction just
discussed. Water plays a critical role, particularly the amount used. The strength of
concrete increases when less water is used to make concrete. The hydration reaction itself
consumes a specific amount of water. Concrete is actually mixed with more water than is
needed for the hydration reactions. This extra water is added to give concrete sufficient
workability. Flowing concrete is desired to achieve proper filling and composition of the
forms. The water not consumed in the hydration reaction will remain in the
microstructure pore space. These pores make the concrete weaker due to the lack of
strength-forming calcium silicate hydrate bonds. Some pores will remain no matter how
well the concrete has been compacted.

Figure 5: Schematic drawings to demonstrate the relationship between the water/cement

ratio and porosity.
The empty space (porosity) is determined by the water to cement ratio. The relationship
between the water to cement ratio and strength is shown in the graph that follows.

Figure 6: A plot of concrete strength as a function of the water to cement ratio.

Low water to cement ratio leads to high strength but low workability. High water to
cement ratio leads to low strength, but good workability.

The physical characteristics of aggregates are shape, texture, and size. These can
indirectly affect strength because they affect the workability of the concrete. If the
aggregate makes the concrete unworkable, the contractor is likely to add more water
which will weaken the concrete by increasing the water to cement mass ratio.
Time is also an important factor in determining concrete strength. Concrete hardens as
time passes. Why? Remember the hydration reactions get slower and slower as the
tricalcium silicate hydrate forms. It takes a great deal of time (even years!) for all of the
bonds to form which determine concrete's strength. It is common to use a 28-day test to
determine the relative strength of concrete.
Concrete's strength may also be affected by the addition of admixtures. Admixtures are
substances other than the key ingredients or reinforcements which are added during the
mixing process. Some admixtures add fluidity to concrete while requiring less water to be
used. An example of an admixture which affects strength is superplasticizer. This makes
concrete more workable or fluid without adding excess water. A list of some other
admixtures and their functions is given below. Note that not all admixtures increase
concrete strength. The selection and use of an admixture are based on the need of the
concrete user.
SOME ADMIXTURES AND FUNCTIONS
TYPE

FUNCTION

AIR ENTRAINING

improves durability, workability, reduces bleeding,

reduces freezing/thawing problems (e.g. special
detergents)

SUPERPLASTICIZERS

increase strength by decreasing water needed for

workable concrete (e.g. special polymers)

RETARDING

delays setting time, more long term strength, offsets

adverse high temp. weather (e.g. sugar )

ACCELERATING

speeds setting time, more early strength, offsets adverse

low temp. weather (e.g. calcium chloride)

MINERAL
ADMIXTURES

improves workability, plasticity, strength (e.g. fly ash)

PIGMENT

adds color (e.g. metal oxides)

Table 3: A table of admixtures and their functions.

Durability is a very important concern in using concrete for a given application. Concrete
provides good performance through the service life of the structure when concrete is
mixed properly and care is taken in curing it. Good concrete can have an infinite life span
under the right conditions. Water, although important for concrete hydration and
hardening, can also play a role in decreased durability once the structure is built. This is
because water can transport harmful chemicals to the interior of the concrete leading to

various forms of deterioration. Such deterioration ultimately adds costs due to

maintenance and repair of the concrete structure. The contractor should be able to
account for environmental factors and produce a durable concrete structure if these
factors are considered when building concrete structures.

Concrete Summary
Concrete is everywhere. Take a moment and think about all the concrete encounters you
have had in the last 24 hours. All of these concrete structures are created from a mixture
of cement and water with added aggregate. It is important to distinguish between cement
and concrete as they are not the same. Cement is used to make concrete!
(cement + water) + aggregate = concrete
Cement is made by combining a mixture of limestone and clay in a kiln at 1450[[ring]] C.
The product is an intimate mixture of compounds collectively called clinker. This clinker
is finely ground into the powder form. The raw materials used to make cement are
compounds containing some of the earth's most abundant elements, such as calcium,
silicon, aluminum, oxygen, and iron.
Water is a key reactant in cement hydration. The incorporation of water into a substance
is known as hydration. Water and cement initially form a cement paste that begins to react
and harden (set). This paste binds the aggregate particles through the chemical process of
hydration. In the hydration of cement, chemical changes occur slowly, eventually creating
new crystalline products, heat evolution, and other measurable signs.
cement + water = hardened cement paste
The properties of this hardened cement paste, called binder, control the properties of the
concrete. It is the inclusion of water (hydration) into the product that causes concrete to
set, stiffen, and become hard. Once set, concrete continues to harden (cure) and become
stronger for a long period of time, often up to several years.
The strength of the concrete is related to the water to cement mass ratio and the curing
conditions. A high water to cement mass ratio yields a low strength concrete. This is due
to the increase in porosity (space between particles) that is created with the hydration
process. Most concrete is made with a water to cement mass ratio ranging from 0.35 to
0.6.
Aggregate is the solid particles that are bound together by the cement paste to create the
synthetic rock known as concrete. Aggregates can be fine, such as sand, or coarse, such
as gravel. The relative amounts of each type and the sizes of each type of aggregate
determines the physical properties of the concrete.

sand + cement paste = mortar

mortar + gravel = concrete
Sometimes other materials are incorporated into the batch of concrete to create specific
characteristics. These additives are called admixtures. Admixtures are used to: alter the
fluidity (plasticity) of the cement paste; increase (accelerate) or decrease (retard) the
setting time; increase strength (both bending and compression); or to extend the life of a
structure. The making of concrete is a very complex process involving both chemical and
physical changes. It is a material of great importance in our lives.
Next Topic: References
Concrete Table of Contents
MAST Home Page

[Journal Home Page] [Search the Journals] [Table of Contents] [PDF version of this
article] [Download to Citation Manager]
J. Phys. Chem. B, 109 (30), 14727 -14734, 2005. 10.1021/jp050237n S10895647(05)00237-3
Web Release Date: July 7, 2005
Copyright 2005 American Chemical Society

Hydration Process of Cement in the Presence of a

Cellulosic Additive. A Calorimetric Investigation
Francesca Ridi, Emiliano Fratini, Francesca Mannelli, and Piero Baglioni*
Department of Chemistry and Consorzio Interuniversitario per lo Sviluppo dei Sistemi a
Grande Interfase (CSGI), University of Florence, Via della Lastruccia 3-Sesto
Fiorentino, I-50019 Florence, Italy
Received: January 13, 2005
In Final Form: May 18, 2005
Abstract:
In the cement industry, the extrusion technique is used to produce flat shapes with
improved resistance to compression. Extrusion is a plastic-forming process that consists
of forcing a highly viscous plastic mixture through a shaped die. The material should be
fluid enough to be mixed and to pass through the die, and on the other hand, the extruded
specimen should be stiff enough to be handled without changing in shape or cracking.
These characteristics are industrially obtained by adding cellulosic polymers to the
mixture. The aim of this work is to understand the action mechanism of these additives
on the major pure phases constituting a typical Portland cement: tricalcium silicate (C3S),
dicalcium silicate (C2S), tricalcium aluminate (C3A), and tetracalcium iron-aluminate
(C4AF). In particular, a methylhydroxyethyl cellulose (MHEC) was selected from the
best-performing polymers for further study. The effect of this additive on the hydration
kinetics (rate constants, activation energies, and diffusional constants) was evaluated by
means of differential scanning calorimetry (DSC) while the hydration products were
studied by using thermogravimetry-differential thermal analysis (TG-DTA), X-ray
diffraction (XRD), and scanning electron microscopy (SEM). MHEC addition in calcium
silicate pastes produces an increase in the induction time without affecting the nucleationand-growth period. A less dense CSH gel was deduced from the diffusional constants in
the presence of MHEC. Moreover, CSH laminar features and poorly structured hydrates
were noted during the first hours of hydration. In the case of the aluminous phases, the

additive inhibits the growth of stable cubic hydrated phases (C3AH6), with the advantage
of the metastable hexagonal phases being formed in the earliest minutes of hydration.
Introduction
Anhydrous Portland cement is a rather complex material mainly constituted by four
phases: tricalcium silicate, Ca3SiO5, dicalcium silicate, Ca2SiO4, tricalcium aluminate,
Ca3Al2O6, and tetracalcium iron-aluminate, Ca4Al2Fe2O10. According to the widely
accepted cement-chemical notation, CaO, SiO2, and H2O become C, S, and H,
respectively, and the single phases can be represented shortly as C3S, C2S, C3A, and
C4AF. Calcium silicates (C3S and C2S) react in water to give the calcium silicate hydrate,
CSH, amorphous gel1 and the crystalline calcium hydroxide, CH, referred to as
portlandite, which has a well-known hexagonal structure. The hydration reaction of C3S
is completed after 1 month, while C2S reacts more slowly and takes more than a year to
be fully hydrated. Tricalcium aluminate and tetracalcium iron-aluminate react with water
to give several metastable (can) phases that successively convert into a cubic structure,
represented by the general formula C3(AF)H6. This structural change is well noticeable
below 30 C, while at higher temperature the conversion is so fast that only the cubic
phase is produced. As a simple approximation, the reaction of tricalcium aluminate with
water may be represented by the equation C3A + 6H C3AH6.2 This product comes from
the rearrangement of more complex hydrates (e.g. C4AH13 and C2AH8), formed as
platelets with hexagonal symmetry,3 or from an initial amorphous gel,4 depending on the
external conditions. The hydration products formed by the ferrite solid solution, C4AF,
are usually described as being similar to those formed by C3A, with Fe3+ partly exchanged
with Al3+.3 The aluminates hydration reaction is faster than that of the silicates, and it is
concluded in a few hours.
Recently, the application of quasi-elastic neutron scattering to the investigation of the
cement setting process has been reported for the hydration of the four different phases.5-7
The time evolution of the translation dynamics of the nonreacted water has been
characterized in detail through the exponent and the average relaxation time in the
framework of the "relaxing cage model". In particular, both parameters confirm a rapid
water confinement in the case of aluminate pastes (C3A and C4AF), while, in the silicates'
case (C3S and C2S), the confinement process is slower, in agreement with the different
kinetic rates and developing hydration products.
The use of natural polymers as proteins or polysaccharides to improve building material
characteristics and performance can be dated back to ancient times. Currently, synthetic
polymers are used in the cement industry to change the cement characteristics to some
extent, in order, for example, to increase the workability and fluidity or to adjust the air
entrainment and the water retention.
The continuous extrusion of cement-based mixtures is one of the actual methods to
produce sheets, bricks, spacers, sewer pipes, and similar objects. It is considered one of
the most promising applications to obtain new cementitious materials, as considerably

less energy is used during continuous extrusion than during production of conventional
ceramic materials.
Extrusion is a processing technique that has been shown to impart high performance
characteristics to fiber-reinforced cementitious materials.8 Basically, a highly viscous,
plasticlike mixture is forced through a die, a rigid opening of desired cross section.
To make the extrusion process possible on cement pastes, an additive must be added to
confer fluidity and plasticity to the final paste and to enable it to be passed through the
die, extruded, and handled without changing in shape or cracking.
One of the best-performing classes of polymers used in this process is that of
methylhydroxyethyl cellulose (MHEC). In the case of C3S, the effect of MHEC has
already been investigated by us using NMR9 and it has been shown that the cellulosic
additive interacts with the water present inside the paste. Under these conditions, the
hydration process becomes more efficient as the water is gradually released from the
cellulose ether to the calcium silicate.
In previous studies,10,11 we have also shown the possibility of investigating the hydration
of C3S as a function of temperature and in the presence of different superplasticizers
(SPs) by determining the amount of unreacted water (free water index, FWI) vs time and
using a novel differential scanning calorimetry (DSC) method. In fact, the fraction of
reacted C3S, , can be calculated from the DSC curves of freezable water melting, with
being inversely proportional to the experimental enthalpy of fusion of freezable water
present in the C3S/water paste. From the FWI parameter, the kinetic mechanism of the
cement hydration process can be obtained, allowing a complete characterization of the
hydration process. The effect of added polymer is then correlated to the kinetic
parameters obtained for the nucleation and diffusion periods of the hydration process.
In this paper, we used DSC at different temperatures (10, 20, 30, and 40 C) to follow
the hydration kinetics (rate constant, activation energy, and diffusional constant).
Differential thermal analysis (DTA) and X-ray diffraction (XRD) were used to provide
evidence of the hydration products, and scanning electron microscopy (SEM) allowed
underlining the morphological changes on the main four phases constituting an ordinary
Portland cement powder.
Experimental Section
Synthetic tricalcium silicate (C3S), dicalcium silicate (C2S), tricalcium aluminate (C3A),
and tetracalcium iron-aluminate (C4AF) were supplied as a gift by CTG, Italcementi
Group, with BET specific surface areas of 0.44 0.05 m2/g, 0.60 0.05 m2/g, 0.48 0.05
m2/g, and 0.54 0.05 m2/g and particles' median radii of 6.0 m, 11.0 m, 16.4 m, and
12.4 m, respectively.
MHEC was obtained from Bayer with the commercial name Walocel MKX 40 000 PP
0.1, a degree of substitution12 (DS) of 1.75, and a molar degree of substitution (MS) of 2.

A 2% aqueous solution of Walocel MKX 40 000 PP 0.1 at 20

000 mPa s.

C had a viscosity of 40

Samples for calorimetric and diffraction measurements were prepared with a 0.4
water/solid ratio by mass. Before preparing the polymer-added samples, MHEC was
prehydrated and a 3.6% w/w solution was obtained: the silicate or aluminate powder was
mixed with this gel and with pure water in order to produce a paste with a 0.4 water/solid
ratio and with 0.27% of MHEC on powder weight. Water was purified with a Millipore
Organex system (R 18 M cm).
DSC measurements were performed using a Perkin-Elmer calorimeter DSC7, connected
to a personal computer; data were elaborated with Pyris software for Windows, version
3.52. The paste (roughly 40 mg) was transferred in a steel pan (diameter 7.4 mm, capacity
60 L, from Perkin-Elmer) that was sealed with the appropriate cover equipped with a
neoprene O-ring to avoid water leaking. Each sample was maintained in a thermostatic
bath at constant temperature (10, 20, 30, and 40 C). Each measurement was carried out
with the following temperature program: from room temperature down to -30 C at 40
C/min, isothermal regime at -30 C for four minutes, from -30 C to -12 C at 20
C/min, and from -12 C to +14 C at 4 C/min. The isothermal step at -30 C was
performed to ensure that free water would freeze despite possible supercooling effects.
Each DSC curve showed a peak relative to water freezing and a peak due to its melting.
From the enthalpy variation of the water melting, we calculated the FWI according to the
procedure reported in previous works.10,11
In the case of DTA and XRD, the hydration reaction was stopped in order to perform the
investigation, since both techniques are affected by the presence of unreacted water.
Several methods to stop the hydration reaction are reported in the literature4,13-16 and are
all based on the criterion that the unreacted water must be substituted by an organic
solvent (easier to evaporate). According to a comprehensive review about the influence of
the blocking protocol on the hydration mechanism of aluminates,17 we applied the
following procedure that we found to be the best: each sample was washed three times
with ethanol and three times with a mixture of ethanol and ether (1:1 v/v), in order to
remove the nonreacted water. At every washing step, the solvent was kept under stirring
for 3 min and then the liquid part was discarded. The so-obtained wet solid was dried for
6 h at 40 C in a nitrogen atmosphere.
DTA analysis was performed by using the SDT 2960, series Q600, TA instrument. The
investigated temperature range was from 30 to 1000 C with a rate increment of 15
C/min.
XRD spectra were obtained using a D8 Advanced of Bruker AXS. Each spectrum was
obtained by scanning from 10 to 90 with a step in 2 equal to 0.02 , corresponding to
an acquisition time of 1 h per spectrum. The spectra were elaborated using the EVA
program, and the phase analysis was made using the ICDD Powder Diffraction File
Database.

SEM observations were carried out by means of a JEOL-JSM 5600 working at 8-10 kV
of acceleration potential. The preparation of the samples for SEM measurements was
performed according to the procedure already described:10 the samples were prepared
with the same chemical compositions as those of the DSC experiments. Pastes were cured
at 15 C for the desired period and fractured and the inner part was chosen for SEM
observation in order to have a sample representative of the bulk hydration and to avoid
surface effects. The electron contrast was enhanced by coating the samples with a gold
film with a JEOL Jee 4B apparatus.
Results and Discussion
Figures 1 and 2 report the time evolution of the FWI obtained for C3S pastes cured at 10,
20, 30, and 40 C in pure water and in the aqueous MHEC solution, respectively. The
curves have the characteristic shape of those for cement hydration, being composed of
three different periods. FWI slowly decreases until time ti, then a sharp fall corresponding
to the onset of the nucleation-and-growth process is observed. A further change of slope
can be clearly seen, indicating the beginning of the diffusional regime, at time td. The
influence of the storage temperature on the duration of the induction (or "dormant")
period is evident from the ti values reported in Table 1 .

Figure 1 FWI vs time for a C3S paste cured in pure water (w/c = 0.4) at 10,
20, 30, and 40 C. As an example, in the inset, the fittings of the acceleration
and diffusional periods are reported for the 40 C curve.
Figure 2 FWI vs time for a C3S paste cured in water (w/c = 0.4) with MHEC
at 10, 20, 30, and 40 C. As an example, in the inset, the fittings of the
acceleration and diffusional periods are reported for the 30 C curve.

FWI was calculated from each measurement according to the following relationship:10,11

where Hexp corresponds to the area of the peak of water melting, wt is the average
weight fraction of water in the pastes, and Htheor is the theoretical value of the fusion
enthalpy of water in the C3S/water paste. In the previous work, immediately after mixing,
Htheor was found to be approximately 85% of the value of pure water (333 J/g), in
agreement with the influence of the sample's heterogeneity on the enthalpy changes as
determined by DSC.18 For this reason, extracted FWI values were corrected by this factor
and, as a result, each curve starts from unity at the beginning of the hydration reaction.
As previously described,10 the Avrami-Erofeev (A-E) model accounts for the nucleation
process.19-21 According to the A-E law, the variation of (degree of reaction) in time is

where ti is the initial time of the acceleration period, i is the degree of reaction at time ti,
k is the rate constant, and M is a parameter dependent on the dimensionality and the type
of growth of the forming particles.
The previous equation can be linearized and written in the form

In this way, given ti, both k and M can be obtained from a linear least-squares fitting on
the linear region of the plot.
The stoichiometric coefficients involved in the hydration reaction of C3S are known to
vary in time, and these variations should be taken into account to obtain a relation
between FWI and the degree of reaction. However, according to Fuji and Kondo,22 the
hydration reaction can be simplified in the following form:

In this way, the relation between FWI and the degree of reaction, , can be simply written
as22,23

where 12.7 is the ratio between the molecular weights of C3S and H2O and w/c is the
water/cement ratio in weight. Since in our case w/c is kept constant and equal to 0.4, the
relation can be further simplified as

Each experimental curve has been fitted according to the A-E law, and the rate constants
deduced from the fitting are listed in Table 1. The comparison of both the curves and the
data obtained for the C3S/water/MHEC paste with those of the C3S/water paste shows that
the presence of the cellulosic additive during the C3S hydration produces two main
effects. The first effect is a slight delay in the setting of C3S paste: both the induction
times, ti, and the deceleration times, td, are increased when the cellulose ether is added to
the paste, even if in a minor proportion to the superplasticizing additives.11 The second
effect is an improved hydration efficiency through an increased availability of the water
in the hydration reaction, as already evidenced by NMR.9 For example, in the case of the
C3S/water/MHEC paste cured at 20 C, the acceleration period ends when only 20% of
the water remains unreacted (FWI = 0.2), whereas, for the C3S/water paste cured at the
same temperature, the water remains nearly 60% unreacted, meaning that, in the presence
of cellulose ether, the C3S can react with most of the available water to complete the
hydration reaction.

As previously noted,10,11,24 ti values follow an Arrhenius-type behavior and the activation

energies for the induction process have been calculated. The activation energy for the
induction period of the paste containing MHEC is 42 kJ/mol. Since the corresponding
value for the C3S cured in pure water is 38 kJ/mol, we can assert that only a slight effect
on the temperature dependence of the induction time at which the cement hydration
reaction starts is produced when the cellulose ether is added to the system.
The analysis of Table 1 indicates that rate constants, k, are temperature-dependent and, in
particular, increase with temperature. Activation energies can be easily obtained from the
Arrhenius equation by plotting ln(k) vs 1/T:

In the presence of cellulosic polymer, the activation energy for the nucleation and growth
process is essentially equal to that obtained with pure water. A value of 35 kJ/mol is
estimated for the MHEC-added paste, not meaningfully different from the value of 37
kJ/mol for the C3S/water system. This result indicates that cellulose ether, in contrast to
superplasticizers (SPs),11 does not affect the hydration thermodynamics; that is, the
MHEC/C3S interaction is negligible. However, the significant influence of this additive
on the water availability suggests that, due to its highly hydrophilic chemical structure, it
strongly interacts with the aqueous phase. Table 2 summarizes the activation energies
for the induction and the acceleration periods of the two systems.
Furthermore, the analysis of the M value does not show any relevant change in
comparison with the case of C3S hydrated in pure water. According to the classical
Avrami's nucleation-and-growth theory,19-21 this parameter combines the time
dependences of the nucleation-and-growth processes and is therefore related to the
morphological characteristics of the growing crystals through the following relation:23,25

where P is the dimensionality constant for the given kinetic process (it can be equal to 1,
2, or 3, corresponding to the growth of fibers, sheets, or polygonal forms, respectively); S
is related to the rate-limiting growth mechanism (it is 1 for a phase boundary growth or 2
when the growth is controlled by diffusion through the liquid phase); and, finally, Q is a
nucleation rate constant which in its limiting cases may be equal to 0 for no nucleation
dependence (or nucleation sites saturation), whereas Q equal to 1 indicates a constant
nucleation rate.
The M average value is around unity, both for C3S/water and for C3S/water/MHEC
systems, and it is, therefore, consistent with two different morphological characteristics
of the growing hydrates: (i) fibers-phase boundary-no nucleation, P = 1, S = 1, and Q = 0,
also reported in the literature as (1,1,0); (ii) sheets-diffusion-no nucleation, P = 2, S = 2,
and Q = 0, (2,2,0). It is worth remarking that the reported calorimetric approach does not
assess any information in the nanoscale dimension and that P, S, and Q parameters only

refer to the overall morphology of the growing nuclei. As atomic force microscopy
(AFM) recent studies have plainly shown, the particles in the micrometer length scale are
actually aggregates of much smaller platelike units. 26,27 The classical nucleation-andgrowth theory applied on our DSC results does not contradict these recent AFM findings.
The SEM micrographs performed on pastes of C3S cured in pure water reveal a onedimensional microstructure of the CSH growing on the particle surfaces, in full
accordance with the literature.10,25,28,29 The presence of MHEC in the paste produces some
significant modification in the appearance of the sample surfaces. The micrographs in
Figure 3A and B show that no relevant difference between this system and the C3S/water
paste can be detected until 12 h; on the contrary, at higher hydration times (Figure 3C and
D), the fibrils of the typical CSH growing phase are not present in the C3S/water/MHEC
pastes, suggesting that the presence of the additive affects the shape of the hydrated CSH.
In the case of the SP compounds, this effect has been ascribed to the direct interaction of
the additives with the C3S surface.11 In a recent article,9 NMR relaxation times
(longitudinal and transversal) for C3S/water and C3S/MHEC/water systems were
discussed and compared with the relaxation times of the MHEC/water mixture. T1 and T2
trends provided evidence of the strong interaction between the water and the cellulosic
polymer. In particular, values of T1 = 950 ms and T2 = 4 ms have been found at the initial
stage of the reaction in the C3S/water paste and are reported in the literature as indicative
of water constrained in capillary pores. The induction period of C3S/water/MHEC is
characterized by T1 = 1.18 s and T2 = 6 ms, essentially the same as those for water
entrapped in MHEC alone (T1 = 1.23 s and T2 = 5.75 ms). This result supports the
hypothesis of a strong interaction between the water and the cellulosic molecules inside
cementitious pastes. It is also evident that this interaction significantly modifies the CSH
morphology. The presence of the cellulosic polymer also changes the appearance of fully
hydrated C3S pastes, as evidenced by the micrographs obtained on samples cured for 1
month (Figure 3E and F) and 3 months (Figure 3G and H): the samples containing
MHEC are much more compact with respect to standard pastes.

Figure 3 SEM micrographs of C3S pastes (w/c = 0.4): (A) bar = 2 m; (B) bar
= 2 m; (C) bar = 1 m; (D) bar = 2 m; (E) bar = 10 m; (F) bar = 10 m;
(G) bar = 5 m; (H) bar = 5 m.

When all the external part of C3S particles has reacted and CSH can no longer be formed
on the surface, the hydration continues toward the interior of the grains. Under these
conditions, the reaction rate is controlled by the diffusion of water across the CSH gel
phase. The relation between the nonreacted water fraction (FWI) and time is the
following:22-24

where Kd is the diffusional constant, td is the initial time of the diffusional period, FWId is
the value of FWI relative to td, and R is the median radius of the particles.
For each curve, data with time greater than td (see Table 1) were considered to follow eq
9. It is evident that the diffusional constant Kd (listed in Table 1) can be calculated from
the slope of the linear plot FWI1/3 vs (t - td)1/2. The values obtained for pastes of C3S in
pure water are very similar to those reported in a previous work with neutron scattering,
utilizing the same w/c ratio and similar surface area of the powder,24 further confirming
the validity of the method proposed here. Several studies showed Kd values to be
independent of temperature.10,24 This is attributed to the competition between two
different contributions, the thermal activation of the diffusional process and the higher
density of the gel phase at higher temperatures.10,24,30,31 In fact, the diffusional constants
would tend to increase with temperature because of the first contribution; however, the
higher the temperature, the more impervious the "walk" of molecules across the medium,
with a consequent decrease of diffusional constants.
As already noted,11 the decreasing trend of the Kd observed for C3S hydration in pure
water, as the temperature increases, indicates that the structure of the hydrated phase
becomes more impervious, hindering the water diffusion through it. When C3S is
hydrated in the presence of MHEC, the trend of the diffusional constants is exactly the
opposite. Kd has been found to increase with the temperature, confirming the controlled
release effect of MHEC on the hydration and indicating that the ordinary thermal
activation is not affected by the disorder of the CSH matrix.
Dicalcium silicate, C2S, is characterized by an extremely slow hydration process. It takes
more than one year to complete the reaction. We acquired all of the kinetics for the
hydration of this phase in pure water and in the presence of MHEC at 20 C. The results
are shown in Figure 4.

Figure 4 FWI vs time for a C2S paste cured in water (w/c = 0.4) with ( ) and
without ( ) MHEC at 20 C.

The shapes of the two curves (C2S with and without additive) are similar, but MHEC is
found to have a slightly retardant effect. Because of the slowness of the reaction, the C2S
system shows two different nucleation-and-growth periods, as it is evident from the two
different slopes in the plot of the A-E law (Figure 5). For this reason, two different
fittings have been applied to each curve. The end of the induction period occurs when 5%
of the available water has reacted: this point marks, therefore, the ti time relative to the
first nucleation-and-growth period. The point where the slope changes indicates the initial

time of the second nucleation-and-growth stage. The results of the fittings are
summarized in Table 3 .

Figure 5 Avrami-Erofeev fittings for the nucleation-and-growth periods of

C2S pastes cured in water (w/c = 0.4) with ( ) and without ( ) MHEC at 20
C.

Thermogravimetry-differential thermal analysis (TG-DTA) and XRD measurements (data

not shown) performed on C3S and C2S powders in pure water and added with MHEC
further confirm the previously described effect of the cellulose ether on the hydration
kinetics. The formation of CSH and Ca(OH)2 (whose decompositions occur at 120 C
and 450 C, respectively) is detected in both the silicate pastes, and the presence of the
additive does not change the chemical nature of the hydration products: nevertheless, a
slight retardation is recorded when the MHEC is added to the mix.
The attempt to obtain reproducible curves by means of DSC for the C3A hydration failed.
This was expected considering the extremely high hydration heat developed upon the
contact of this phase with water (850 J/g) that makes the control of the temperature
during the sample preparation very hard. In contrast to the C3A case, the time evolution of
FWI in the C4AF/water and C4AF/water/MHEC systems is reproducible, due to the lower
hydration heat of C4AF (420 J/g). However, as already specified, the nature of the
hydrated phases for the aluminates is temperature-dependent, and the hydration behavior
of these phases with the temperature is anomalous with respect to C3S. In particular, the
hydration rate increases as the temperature passes from 10 to 30 C (data not shown); a
further increment of the curing temperature up to 40 C produces a decline in the
hydration rate, in complete agreement with the literature.32,33 Moreover, the FWI value
obtained after the first minutes of hydration is extremely low. During the first 15 min,
approximately 50% of the water reacts with the solid phase, becoming immediately
undetectable with our method. Because of these reasons, it seems obvious that the DSC
approach is partly appropriate to inspect the cellulose ether influence on C3A and C4AF
hydration, since the earliest minutes of the reaction are decisive.
We utilized TG-DTA measurements to investigate the MHEC influence on the aluminate
phases. We examined samples of C3A and C4AF powders mixed at 15 C with water and
with and without MHEC, and we observed the time evolution of the thermal effects
produced from the decomposition of the hydrated phases. Particular attention was paid to
the early stages of the process when most of the water is consumed in the heterogeneous
reaction with the rapid formation of hydrated phases.
As an evidence of the extremely high reaction rate of the aluminate phases, it has to be
noted that immediately after the mixing of both C3A and C4AF with water some reaction
products are formed (Figures 6 and 7). Just after 5 min, the C3A/water system presents
two peaks at 128 C and 280 C (Figure 6A): with the progress of the hydration, the

intensity of this latter peak increases, with a corresponding progressive shift toward
higher temperature, up to 316 C, after 4 days of curing. Previous studies performed on
aluminoferrites34 and on metakaolin35 assign the decomposition occurring around 300 C
to stable cubic hydrogarnet phases. On the other hand, the XRD analysis performed on
these samples effectively evidences the presence of C3AH6 (the cubic hydrogarnet
product relative to the tricalcium aluminate hydration), together with some unreacted
C3A. No other XRD pattern is detectable, confirming the amorphous nature of the
hydrated product responsible for the DTA peak at 128 C.4 The peak at 450 C,
registered on the sample cured for 4 days, is due to the loss of a remaining mole of crystal
water from the hydrogarnet phase.34

Figure 6 Temperature dependence of the heat flux for (A) C3A/water and (B)
C3A/water/MHEC pastes after different times of hydration: 5 min ( ), 15 min
( ), 30 min ( ), 1 h ( ), and 4 days ( ). The curve of dry C3A has been
reported as reference ( ). An increasing negative offset has been added as a
function of time for display purpose only.

Figure 7 Temperature dependence of the heat flux for (A) C4AF/water and (B)
C4AF/water/MHEC pastes after different times of hydration: 5 min ( ), 15
min ( ), 30 min ( ), 1 h ( ), and 4 days ( ). The curve of dry C4AF has been
reported as reference ( ). An increasing negative offset has been added as a
function of time for display purpose only.

When the MHEC is added to the paste (Figure 6B), a relevant inhibition in the formation
of the cubic stable phase occurs in the first minutes of the reaction, as testified by the
absence of the peak at 300 C after 5 and 15 min from the mixing. These samples
present endothermal features at 145, 230, and 260 C. According to the XRD spectra, the
hydrated products formed under these conditions are not crystalline, since no additional
pattern is present apart from that of the anhydrous tricalcium aluminate. The presence of

the cellulose ether during the early stages of C3A hydration stabilizes the amorphous gel,
precursor of the stable C3AH6. At hydration times over 30 min, the thermograms are
identical to those obtained for the C3A/water system.
The thermograms for the C4AF/water and C4AF/water/MHEC systems show
decompositions approximately at the same temperatures previously reported for the C3A
hydration, indicating that the reaction mechanism of the two aluminate phases is the
same. When C4AF is incubated with pure water (Figure 7A), the formation of the cubic
product (C3(AF)H6) is retarded with respect to the case of the tricalcium aluminate
system. On the contrary, the phases whose water loss occurs at 150 and 230 C are
present in a higher amount than those in the C3A/water paste. Again, the XRD confirms
the presence of the cubic hydrogarnet product and of the amorphous gel.
MHEC influences the early stages of the C4AF hydration in a striking way: it totally
inhibits the growth of the stable C3(AF)H6 during the first hour after mixing. Only after 4
days of hydration are the decomposition peaks of the C3(AF)H6 (at 300 and 450 C)
recognizable (Figure 7B).
Figure 8 shows some micrographs obtained on the C3A/water and C3A/water/MHEC
samples. C3A cured in pure water for 5 min (Figure 8A) shows a compact and featureless
structure. When the cellulose ether is added to the mix, the irregular foils, typical of the
amorphous gel, appear. 4 When both the polymer-free and the polymer-added systems are
cured for 1 h or more, their morphologies are dominated by the cubic phase, C3AH6.
Figure 8C and D shows the structures after 4 days of hydration.

Figure 8 SEM micrographs of C3A pastes (w/c = 0.4): bar = 5 m.

Similarly, the bulk of the C4AF/water paste cured for 5 min appears as a featureless
structure with some sporadic regions of irregular foils (Figure 9A). The cellulose ether
produces a considerable stabilization of this acicular structure in the early stages of the
C4AF hydration: the samples cured both for 5 min (Figure 9B) and for 1 h (not shown)
are characterized by these features. After 4 days, only the cubic phase is visible in both
the systems (Figure 9C and D).

Figure 9 SEM micrographs of C4AF pastes (w/c = 0.4): (A) bar = 2 m; (B)
bar = 2 m; (C) bar = 5 m; (D) bar = 2 m.

Conclusions
In the present work, the hydration of the main four phases constituting an ordinary
Portland cement has been investigated. DSC allows us to perform an "in situ" analysis of
the hydration kinetics of Ca3SiO5 (C3S) and Ca2SiO4 (C2S) phases in the presence of
methylhydroxyethyl cellulose (MHEC). C3S/water and C3S/water/MHEC systems have
been monitored at different temperatures (10, 20, 30, and 40 C), in order to perform a
complete kinetic analysis: the activation energies for the induction period and for the
nucleation-and-growth process of the hydrated phase, the M parameter (related to the
morphological characteristics of the growing crystals), and the diffusional constants have
been obtained for these systems. The presence of cellulose ether does not significantly
alter the activation energy for the acceleration period, and also, the M parameter remains
roughly constant. These results indicate that no modification in the C3S hydration
mechanism occurs in the presence of cellulose ether. However, FWI evolution highlights
a strong increment of the water availability during the acceleration period, as a
consequence of MHEC presence: the additive, due to its highly hydrophilic chemical
structure, interacts with the aqueous phase, making it homogeneously distributed over the
paste. As a consequence, the H2O does not need to diffuse to reach the anhydrous C3S and
is available to the hydration in a shorter period of time. We also studied the hydration
reaction at 20 C of the C2S phase with and without MHEC: in these systems, two steps
in the formation of the hydrated phase can be identified. The rate constants of these two
nucleation-and-growth periods have been calculated.
The evolution of the hydration reaction of Ca3Al2O6 (C3A) and Ca4Al2Fe2O10 (C4AF)
phases was monitored by differential thermal analysis (DTA): the MHEC was shown to
stabilize the formation of the metastable amorphous aluminate hydrated gel in the early
minutes of the reaction, with further inhibition of the formation of the stable phase, the
cubic hydrogarnet C3(AF)H6. SEM was used to examine the morphological modifications
occurring on the main four phases because of the presence of cellulose ether.
Acknowledgment
The authors thank Dr. L. Cassar, R. Alfani, G. Guerrini, M. Biagini (CTG-Italcementi
Group), and L. Dei (CSGI) for their invaluable comments and discussions. Financial
support from CTG-Italcementi Group, Ministero dell'Istruzione, Universit e della
Ricerca Scientifica (MIUR, Grant PRIN-2003), and Consorzio Interuniversitario per lo
Sviluppo dei Sistemi a Grande Interfase, CSGI, is gratefully acknowledged.
* To whom all correspondence should be addressed. Phone: +39 055 457-3033. Fax: +39
055 457-3032. E-mail: piero.baglioni@unifi.it. Website: www.csgi.unifi.it.
1. Xu, Z.; Viehland, D. Phys. Rev. Lett. 1996, 77, 952-955. [ChemPort] [CrossRef]
2. Soroka, I. Portland Cement Paste and Concrete; The MacMillan Press LTD: 1979.

3. Bye, G. C. Portland cement. Composition, production and properties, 2nd ed.; Thomas
Telford: London, 1999.
4. Breval, E. Cem. Concr. Res. 1976, 6, 129-138. [ChemPort] [CrossRef]
5. Fratini, E.; Chen, S. H.; Baglioni, P.; Bellissent-Funel, M. C. J. Phys. Chem. B 2002,
106, 158-166.[Full text - ACS] [ChemPort]
6. Fratini, E.; Chen, S. H.; Baglioni, P.; Cook, J. C.; Copley, J. R. D. Phys. Rev. E 2002,
65, 010201. [ChemPort] [CrossRef]
7. Fratini, E.; Chen, S. H.; Baglioni, P. J. Phys. Chem. B 2003, 107, 10 057-10 062.
8. Shao, Y.; Marikunte, S.; Shah, S. P. Concr. Int. 1995, 17, 48-52. [ChemPort]
9. Alesiani, M.; Capuani, S.; Giorgi, R.; Maraviglia, B.; Pirazzoli, I.; Ridi, F.; Baglioni, P.
J. Phys. Chem. B 2004, 108, 4869-4874.[Full text - ACS] [ChemPort]
10. Damasceni, A.; Dei, L.; Fratini, E.; Ridi, F.; Chen, S. H.; Baglioni, P. J. Phys. Chem.
B 2002, 106, 11 [Full text - ACS] 572-11 578.
11. Ridi, F.; Dei, L.; Fratini, E.; Chen, S. H.; Baglioni, P. J. Phys. Chem. B 2003, 107,
1056-1061.[Full text - ACS] [ChemPort]
12. DS indicates the average number of etherified hydroxyl groups in a glucose unit,
whereas MS indicates the average number of bonded hydroxyalkyl groups per monomer
unit. While the maximum obtainable DS is 3 (number of hydroxyl groups in a glucose
unit), MS has theoretically no limit, since ethylene oxide can form side chains during the
reaction.
13. Smith, A.; Chotard, T.; Gimet-Breart, N.; Fargeot, D. J. Eur. Ceram. Soc. 2002, 22,
2261-2268. [ChemPort] [CrossRef]
14. Guirado, F.; Gali', S.; Chincon, J. S. Cem. Concr. Res. 1998, 28, 381-390. [ChemPort]
[CrossRef]
15. Stoch, A.; Zdaniewicz, M.; Paluszkiewicz, C. J. Mol. Struct. 1999, 511-512, 319-325.
16. Georgescu, M.; Puri, A.; Coarna, M.; Voicu, G.; Voinitchi, D. Cem. Concr. Res. 2002,
32, 1269-1275. [ChemPort] [CrossRef]
17. Bachiorrini, A.; Guilhot, B. Cem. Concr. Res. 1982, 12, 559-567. [ChemPort]
[CrossRef]
18. Wendlandt, W. W. Thermal analysis, 3rd ed.; Wiley-Interscience: New York, 1986.

19. Avrami, M. J. Chem. Phys. 1939, 7, 1103-1112. [ChemPort]

20. Avrami, M. J. Chem. Phys. 1940, 8, 212-224. [ChemPort]
21. Avrami, M. J. Chem. Phys. 1941, 9, 177-184. [ChemPort]
22. Fujii, K.; Kondo, W. J. Am. Ceram. Soc. 1974, 57, 492-497. [ChemPort]
23. Berliner, R.; Popovici, M.; Herwig, K. W.; Berliner, M.; Jennings, H. M.; Thomas, J.
J. Cem. Concr. Res. 1998, 28, 231-243. [ChemPort] [CrossRef]
24. FitzGerald, S. A.; Neumann, D. A.; Rush, J. J.; Bentz, D. P.; Livingston, R. A. Chem.
Mater. 1998, 10, 397-402.[Full text - ACS] [ChemPort]
25. Brown, P. W.; Pommersheim, J.; Frohnsdorff, G. Cem. Concr. Res. 1985, 15, 35-41.
[ChemPort] [CrossRef]
26. Garrault-Gauffinet, S.; Nonat, A. J. Cryst. Growth 1999, 200, 565-574. [ChemPort]
[CrossRef]
27. Nonat, A. Cem. Concr. Res. 2004, 34, 1521-1528. [ChemPort] [CrossRef]
28. Double, D. D.; Hellawell, A. Nature 1976, 261, 486-488. [ChemPort]
29. Double, D. D.; Hellawell, A. Nature 1977, 237, 82-90. [ChemPort]
30. Escalante-Garca, J. I.; Sharp, J. H. Cem. Concr. Res. 1998, 28, 1245-1257.
[ChemPort] [CrossRef]
31. Massazza, F.; Daimon, M. Presented at 9th International Congress on the Chemistry
of Cement, New Delhi, India, 1992.
32. Capmas, A. In Calcium Aluminate Cements; Mangabhai, R. J., Ed.; E&FN Spon:
London, 1990.
33. Bushnell-Watson, S. M.; Sharp, J. H. Cem. Concr. Res. 1990, 20, 677-686.
[ChemPort] [CrossRef]
34. Csizmadia, J.; Balazs, G.; Tamas, F. D. Cem. Concr. Res. 2001, 31, 577-588.
[ChemPort] [CrossRef]
35. Rojas, M. F.; Sanchez de Rojas, M. I. Cem. Concr. Res. 2003, 33, 643-649.
[ChemPort] [CrossRef]

Table 1: Kinetic Parameters of the Nucleation-and-Growth Process and of the Diffusional

Period for C3S Hydration in Pure Water and with MHEC at the Four Investigated
Temperatures
nucleation-and-growth

diffusion-limited

T ( C)

ti (h)

k (h-1)

td (h)

Kd 1015 (m2/h)

water

10

0.0113

1.3

38

13.7

water

20

0.0234

1.1

32

4.33

water

30

0.0271

1.0

18

3.72

water

40

1.5

0.0579

1.1

9.5

2.99

MHEC

10

32

0.0296

1.1

69.5

4.55

MHEC

20

15

0.0664

1.2

47

6.42

MHEC

30

0.0744

1.2

37.5

8.67

MHEC

40

0.1390

0.9

24

13.0

Table 2: Activation Energies for the Induction Period and for the Acceleratory Stage of the
Hydration of C3S in Pure Water and in the Presence of MHEC
Ei (kJ/mol)

Ea (kJ/mol)

water

38

37

MHEC

42

35

Table 3: Parameters for the Nucleation-and-Growth Processes of C2S Hydration in Pure

Water and with MHEC at 20 C
ti (h)

k (h-1)

water-1

0.0071

0.7

water-2

743

0.0036

0.8

MHEC-1

22.5

0.0025

0.8

MHEC-2

743

0.00063

1.0

Part 4. Hydration of Portland Cement

Chemistry of Hydration - It is assumed that each compound hydrates
independently of others in Portland cement. This is not completely true
because interaction between hydrating compounds will affect the mix.
Calcium Silicates - The hydration reaction of the two calcium silicates, which
make up the largest percent of Portland cement, are similar.
The principle products are: 1) calcium silicate hydrate, poorly crystalline
material of extremely small particle size and 2) calcium hydroxide, a
crystalline material. The reaction can be measured by the rate of heat
generation.
Stage 1 - Rapid heat generation (15 min.) -- on mixing with water, calcium
and hydroxide ions are released from the surface of the C3S; pH rises to a
very alkaline solution. When the calcium and hydroxide reach critical
concentrations, crystallization of CH and C-S-H begins. Early chemical
reactions are temperature dependent.

Stage 2 - Dormant period - causes cement to remain plastic (2-4 hours). The
reaction slows. CH crystallizes from the solution, C-S-H develops on the
surface of the C3S and forms a coating. As the thickness increases, the time it
takes water to penetrate the coating increases, thus the rate of reaction
becomes diffusion controlled. C<small><small>2</small></small>S hydrates
at a slower rate because it is a less reactive compound.
Stage 3 - Acceleration period - Critical concentration of ions is reached and
silicate hydrates rapidly, maximum rate occurs at this stage. Final set has
passed and early hardening begins (4-8 hours).
Stage 4 - Deceleration - rate of reaction slows; completely diffusion
dependent reaction.
Stage 5 - Steady state - constant rate of reaction (12-24 hours). Temperature
has little effect on hydration at this point.
Tricalcium Aluminate - Hydration of C3A occurs with sulfate ions supplied by
dissolved gypsum. The result of the reaction is calcium sulfoaluminate hydrate,
called "ettringite" after a naturally occurring mineral.
If the supply of sulfate from the gypsum is exhausted before the C3A is
completely hydrated, a second reaction can occur. The product of this
reaction is monosulfoaluminate. This reaction may occur before the
formation of the ettringite if the reaction of C3A and the sulfate ions is faster
than the gypsum will allow.
The ettringite decreases the reaction by forming a diffusion coating around
the C3A similar to the reaction of C3S. The coating can be broken down by the
conversion to monosulfoaluminate.
If the monosulfoaluminate is exposed to another source of sulfate ions, the a
new reaction will occur forming more ettringite. This new formation causes
volume to increase and leads to tensile cracking. This tendency is the basis
for sulfate attack of Portland cements.
In the absence of sulfates, C3A reacts with water to form two unstable
calcium hydrates which later convert to hydrogarnet. This is the same
process found in HAC. A pure C3A paste will not develop significant strength.
Ferrite Phase (C4AF) forms the same hydration products as C3A, with or without
gypsum. The reaction is slow and is decreased further by gypsum. If the iron
oxide content is increased, the reaction is slower.
Experience has shown cements low in C3A and high in C4AF are sulfate

resistant. The conversion from ettringite to monosulfoaluminate is inhabited

by the presence of the iron component.
** The rate of hydration is on the order of C3A > C3S > C4AF > C2S. Reactions
for even identical compounds may vary due to: 1) fineness, 2) rate of cooling
of clinker, and 3) impurities.

Properties of the Hydration Products

Some general comments on the properties of hydration products affecting the
overall behavior of the cement.
C-S-H, calcium silicate hydrate -- very poor crystallinity; the exact chemical
compound is variable. The ratio of C/S varies between 1.5 and 2.0 and depends
on many factors; temperature, w/c ratio, impurities, etc. Likewise, measures
of the water content vary considerably.
Because of the poor crystallinity, C-S-H develops very small irregular particles
and consequently a very high surface area. In general, the surface area of the
hydrated cement is about 1000 times larger than the unhdyrated cement.
Therefore, the increase in surface area greatly influences physical properties of
the C-S-H hydrate.
Considerable work has been done in modeling the structural components of CS-H, with much disagreement among scientists. C-S-H is considered a layer
structure composed of calcium silicate sheets randomly connected by strong
ionic-covalent bonds. The remainder of the interlayer space is classified as: 1)
capillary pores, relatively large openings where water can form menisci; 2)
micropore, smaller spaces where water cannot form menisci. The water forces
the layers apart by exerting a disjoining pressure. This pressure decreases with
lower water content; 3) interlayer space, layers are close enough that the
trapped water bonds the sheets together by van der Waal forces.
There are three accepted models for the C-S-H structure:
1. Powers-Brunauer -- a chemical structure model based on the layered
claylike configuration. The sheets are randomly arranged and contain
absorbed water on their surfaces. Water can penetrate the interlayer,
micropore region. Interlayer water can be permanently removed with
strong drying.
2. Feldman and Sereda -- a model composed of a completely random array
of single layers forming irregular interlayer space. Water can move in
and out of the interlayer space even after drying.
3. Munich -- a physical model where C-S-H is considered a threedimensional arrangement of colloidal particles. The chemical nature of

the model is secondary. Van der Walls' forces bind surfaces together, but
strong covalent bonds are more common. Water is attracted to the solid
particles resulting in disjoining pressures which reduce particle
interaction.
*** In all these models, water affects the structure of the C-S-H. As water is
removed from capillaries, a resultant compressive stress is induced. Loss of
interlayer water affects particle or sheet bonding.
Calcium Hydroxide -- a well understood hexagonal crystalline material.
Crystals are much larger than C-S-H particles and are sometimes visible to the
naked eye.
Calcium Sulfoaluminate (ettringite) -- These hexagonally-shaped prism
crystals are considerably longer than CH crystals. Large clusters of ettringite
needles may be visible in concrete affected by sulfate attack.
Monosulfoaluminate tends to form very thin, hexagonal plates.

Microstructure of Hydrated Cement

Pastes
The development of cement microstructure relates to the five chemical stages
described earlier in this chapter.
C-S-H -- the largest component of the cement paste (50-70%) and is the most
important component in the hydration process. The amount of C-S-H coating on
a C3S grain is very small during stage 2 of hydration and increases rapidly in
stage 3. The spines of the forming C-S-H radiate outward from each grain with
the bulk of the material below the spines. As the C-S-H hydrates further, the
coating thickness grows forcing the outward spines of adjacent particles to
interlock to form solid bonds. As hydration continues the intermeshed spines
contribute to an increase in the undercoating of C-S-H growth. The effect is to
bond the cement grains together with the C-S-H coating.
CH -- constitutes 20-25% of the cement volume. In the acceleration stage, CH
grows in the capillary pore space. CH will only grow in free space; on
encountering another CH crystal it will grow in another direction; also it will
grow completely around a hydrating cement grain. The latter effect gives the
CH a larger apparent volume in cement pastes than it would have as a pure
crystal.
Calcium Sulfoaluminate -- a small component of cement pastes (10-15%)
having little effect on microstructure. Young spiny ettringite crystals grow into
capillary space and later convert to flat monosulfoaluminate crystals. There

will be unhydrated residues in the cement paste, mainly caused by calcium

hydroxide, even in very matured hydrated pastes.
Porosity -- a major component of microstructure which will influence paste
properties. Pore size distribution is difficult to measure. Many tests require
drying, which affects the pore structure. There are two classifications of pore
sizes:
1. Capillary pore -- space formed between hydrating gains.
2. Gel pores -- very small spaces in the C-S-H coating. Constitutes the bulk
of porosity in a cement paste.

Properties of Hydrated Cement Pastes

Hydration products have lower specific gravities and larger specific volumes
than their parent cement compounds. Therefore, every hydration reaction is
accompanied by an increase in solid volume.
Calcium Silicates -- hydration of these materials is not accompanied by an
increase in volume. Recall, these crystals will only occupy free space. If this
space is filled, the growth or hydration will stop.
Calcium Aluminate -- The hydration product of this material (ettringite) will
continue to form even when a solid surface is encountered. Since there is no
volume in which the crystal can grow, internal pressures develop.
Volume change is directly related to porosity. It is possible to calculate pore
space by measuring the loss of evaporable water and nonevaporable water. The
evaporable water describes water held in capillary and gel pores. This amount
can be determined by oven drying a sample. Nonevaporable water is a measure
involving the microstructure of the hydration product and is obtained from a
paste heated to very high temperatures (1000 C0). T.C. Powers developed
several empirical relationships for degree of hydration based on the amount
the two types of water described above.
wn = 0.24a g/g of original cement
where a = degree of hydration and wn = nonevaporable water
wg = 0.18a g/g of original cement
where wg = gel water or evaporable water

Other relationships for volume of hydration products and porosity are available
(see p. 105). Based on these, a minimum water/cement ratio relationship for
complete hydration can be formed.
wmin = ( wn + wg ) g/g of original cement
(w/c)min = 0.42a
Therefore, for complete hydration, the w/c ratio should not fall below 0.42.
However, complete hydration is not required for high ultimate strength. This
means that paste with low w/c ratios will self-desiccate unless external water
is added. Generally, this is not a problem in the field.

This web site was originally developed by Charles Camp for his CIVL 1101 class.
This site is maintained by the Department of Civil Engineering at the University
of Memphis. Your comments and questions are more than welcome.

components
%

CaO
62.3
64.4

SiO2
21.3
25.0

Al2O3
3.4 5.2

Fe2O3
2.3 4.1

MgO
1.8 3.0

SO3
1.6 3.1

LOI
0.8 1.0

LOI = Loss of ignition = fly ash, metakaolin (cao lanh) , silica fume

Drawing 3 parallel lines with 3 edges of ternary diagram representing overall composition of calcareous
cementitious materials from any point (A) of Portland cement area shows rate of principal compounds in
Portland cement such as CaO, SiO2 and Al2O3. It is recognized that CaO, SiO2 and Al2O3 hold about 63.5%,
23% and 13.5 % respectively. These is average value because