Cement Chemistry




Mixture Design

Fresh Concrete

Hardened Concrete

Dimensional Stability




Hydration of cement

The reaction of cement hydration is exothermic. Measurements using a conduction calorimeter can give the rates of heat evolution at various stages.

A typical heat evolution pattern from cement hydration is presented in Figure 1. There are three characteristic peaks for ordinary Portland cement. The initial heat burst corresponds to the instantaneous high rate of heat evolved when cement is brought into contact with water. This is due to the heat of wetting (Heat of wetting = Surface energy – Energy required for interface creation). Hydration of C3S and C3A also contribute to this peak.

Figure 1: Heat evolution during cement hydration

The initial burst is followed by a slowdown of the heat evolution rate. The rate does not become negative or zero at any stage, implying that although slowly, the reactions do continue. This is termed as the ‘dormant’ or the ‘induction’ period. This period is followed by the main peak of cement hydration, which is associated with the rapid dissolution of C3S to form CSH and CH, and formation of ettringite (AFt) from C3A.

A slowdown of the hydration process beyond the main peak leads to lower rates of heat evolution. A broader peak is associated with the conversion of ettringite to monosulphate (AFm).

The latest calorimeters can also detect an extra endothermic peak in the beginning that corresponds with the dissolution of potassium sulphate (when it is present in the cement).

It is difficult to obtain the correct relationship between heat evolution and temperature unless the system is perfectly insulated. Another problem is the dependence on the water to cement ratio. Water has a much higher specific heat than cement, thus when more water is present, a higher degree of heat will be required to increase the temperature of the system.

Cement contains highly soluble alkali oxides (Na2O and K2O). The dissolution of these compounds is responsible for the high alkalinity (pH 12 – 13) of the pore solution. Thus, the hydration of cement actually takes place in the pore solution, and not in water.

Dormant Period

Various theories have been proposed for the existence of the dormant period. As stated earlier, the rate of heat evolution during this stage is low. The slowdown of the hydration process has been explained using the following ideas:

  1. Formation of an impermeable hydrate layer (CSH) on the surface of the C3S particle precludes the further dissolution of C3S.
  2. The hydrate layer has a lower C/S ratio compared to C3S. As a result Ca2+ is released into the liquid phase (which contains OH-), and a silica rich layer forms on the surface of the C3S particle. This electrical double layer thus formed prevents any reaction to form CSH by impeding the passage of ions.
  3. Liquid phase gets supersaturated with respect to CH. As a result CH starts precipitating and this stops the further dissolution of C3S.

The end of the dormant period can come about in many ways:

  1. The barrier can weaken due to ageing.
  2. Diffusion of ions can occur across the barrier by osmosis.
  3. A gradual weakening of the electrical double layer may occur.
  4. Nucleation of CH can get slowed down when the nuclei start approaching their critical size.

Reactions during hydration

Reactions involving the silicates

2 C3S + 6 H —›C3S2H3 + 3 CH (1)
2 C2S + 4 H —›C3S2H3 + CH (2)

The above reactions are perfectly stoichiometrically balanced. However, C-S-H does not have a well defined stoichiometry. The C/S of C-S-H can vary from 1.5 to 2, and commonly is around 1.8. The main difference in the hydration of the two silicates lies in the amount of CH formed in the reaction. It is evident from the above equations that 3 times as much CH is formed from C3S hydration as in C2S hydration.

C-S-H does not have a definite structure is thus termed as a gel. CH deposits as hexagonal crystals, generally oriented tangentially to pore spaces and aggregates along the longitudinal axis.

Reactions involving the aluminates

In the absence of gypsum, calcium aluminate hydrates form from C3A, resulting in a flash set of the cement paste.

2 C3A + 21 H —›C4AH13 + C2AH8 (3)

C2AH8 is a metastable phase that deposits as hexagonal platelets (similar to CH). Above 30 oC, it is converted to cubic hydragarnet (C3AH6).

In the presence of gypsum, ettringite formation occurs.

C3A + 3 C S H2 + 26 H —›C6A S 3H32 (4)

Ettringite (or the AFt phase) gets deposited as acicular, columnar, hexagonal crystals. The presence of tubular channels in between the columns can lead to high water absorption and swelling by ettringite. This is one of the theories explaining the expansion caused by ettringite formation.

Nearly all the SO42- gets combined to form ettringite in an ordinary Portland cement. If there is still C3A left after this reaction, it can combine with ettringite to form monosulphate (or AFm phase) which has a stoichiometry of C4A S H12-18. If there is sufficient excess C3A, then C4AH13 can also form as a hydration product, and can exist in a solid solution with AFm.

C4AF produces similar hydration products as C3A, with the Al3+ being partly replaced by Fe3+. The final hydration product depends on the availability of lime in the system. In the presence of gypsum, C4AF produces an iron-substituted ettringite. Higher the ratio C4AF/C3A, lower is the conversion of ettringite to monosulphate.

Kinetics of cement hydration

The progress of cement hydration depends on:

  • Rate of dissolution of the involved phases (in the initial stages), and at later stages,
  • Rate of nucleation and crystal growth of hydrates
  • Rate of diffusion of water and dissolved ions through the hydrated materials already formed

The factors affecting the kinetics of hydration are:

  1. The phase composition of cement
  2. The amount and form of gypsum in the cement: Whether gypsum is present in the dihydrate, hemihydrate, or the anhydrite form.
  3. Fineness of cement: Higher the fineness, higher the rate of reaction due to availability of a larger surface area.
  4. w/c of mix: At high w/c, hydration may progress till all of the cement is consumed, while at low w/c the reaction may stop altogether due to lack of water.
  5. Curing conditions: The relative humidity can have major effects on the progress of hydration.
  6. Hydration temperature: Increase in temperature generally causes an increase in the rate of the reaction, although the hydrated structure can be different at different temperatures.
  7. Presence of chemical admixtures: For example, set controllers, and plasticizers.

Stages in Cement Hydration

1. Pre-induction period (first minutes):

  • Rapid dissolution of ionic species (alkali sulphates contribute K+, Na+, and SO42-; CaSO4 dissolves until saturation, contributing Ca2+ and SO42-)
  • C-S-H forms on the surface of dissolving C3S. C/S of C-S-H is lesser than of C3S, thus an increase of the Ca2+ concentration in the liquid phase occurs. Formation of electrical double layer, and the precipitation of CH leads to the dormant period.
  • C3A dissolves, and reacts with SO42- to form AFt, which forms a surface barrier. C4AF also reacts to form AFt.
  • Only very small % of C2S reacts at this stage.

2. Induction (dormant) period (first few hours):

  • CH concentration in the liquid phase reaches a maximum and then starts to decline.
  • The concentration of SO42- remains constant as the amount consumed due to AFt formation is balances by the amount dissolved from gypsum.

3. Acceleration stage (3 – 12 hours after mixing):

  • Nucleation and growth of C-S-H (often termed as the ‘second-stage CSH’) and CH occurs. C2S also starts hydrating substantially.
  • Ca2+ concentration in the liquid phase declines as Ca(OH)2 starts precipitating.
  • SO42- concentration starts to decline with increasing AFt formation, and adsorption of SO42- on C-S-H.

4. Post-acceleration period:

  • Slow down due to decline in non-reacted material, and because the process becomes diffusion controlled.
  • The contribution of C2S increases steadily, leading to a decline in the rate of formation of CH.
  • Consumption of SO42- leads to a conversion of AFt to AFm.

The progress of hydration, both in terms of the unhydrated compounds consumed, as well the hydration products formed, has been presented in Figure 2. In the first few minutes, about 2 – 10 % of C3S hydrates, and a significant fraction is consumed within 28 days. The rate of hydration depends upon the reactivity of alite (i.e. the amount of foreign ions present within the alite structure). With an increase in the amount of SO3, the C3S reaction becomes faster. However, beyond a limit, SO3 can start causing retardation.

The hydration of C2S is a slow process, and does not pick up for many hours. On the other hand, 5 – 25% of C3A reacts in the first few minutes of hydration. The initial reactivity depends on the quantity and quality of alkalis present (K+ increases reactivity, while Na+ decreases it).The reactivity of C4AF is dependent on the A/F of the cement.
The method of grinding cement may also influence the hydration kinetics. Cements ground in high pressure roller mills set faster than in ball mills, because of higher reactivity of C3A and C3S phases, and a lowered rate of decomposition of CaSO4.

Figure 2: Progress of cement hydration

Composition of pore solution

The evolution of pore solution composition for a typical cement (0.6% equivalent Na2O, 3% SO3, 0.5 w/c) is shown in Figure 3. By 1 week, the only ions remaining in appreciable concentration are Na+, K+, and OH-. The concentration of OH- is almost a mirror image of that of SO42-, due to considerations of ionic balance within the pore solution. Ground clinker would typically have a lower ionic concentration in the pore solution due to the absence of SO42-.

Figure 3: Evolution of the pore solution composition in cement paste

Structure of hydrated cement paste

The following micrographs (obtained by Scanning Electron Microscopy) reveal the typical features of a hydrated cement paste.

Figure 4: Fracture surface of a PC mortar showing the gel-like nature of C-S-H

Figure 5: Image of polished surface of a PC mortar; the bright particles are that of unhydrated cement; the grayish background is the C-S-H, while the white rims around the aggregate pieces are deposits of CH


Figure 6: Polished surface of a C3S mortar showing hydrating grains of C3S; the darker shades are C-S-H deposits, while the lighter shades, especially as rims around aggregates are deposits of CH


Figure 7: Low magnification image showing the polished surface of PC concrete; the porosity of the paste and the extent of unhydrated cement are visible

Hydrated cement paste is composed of capillary pores and the hydration product. The pores within the structure of the hydration product are termed ‘gel’ pores. This hydration product includes C-S-H, CH, AFt, AFm, etc. Gel pores are included within the structure of hydrated cement. According to Powers, 1/3 of the pore space is comprised of gel pores, and the rest are capillary pores. The pores inside cement paste contain water (or pore solution), which can be classified into:

  1. Capillary water: Present in voids larger than 50 Ao. Further classified into: (a) free water, the removal of which does not cause any shrinkage strains, and (b) water held by capillary tension in small pores, which causes shrinkage strains on drying.
  2. Adsorbed water: Water adsorbed on the surface of hydration products, primarily C-S-H. Water can be physically adsorbed in many layers, but the drying of farther surfaces can occur at about 30 % relative humidity. Drying of this water is responsible for a lot of shrinkage.
  3. Interlayer water: Water held in between layers of C-S-H. The drying of this water leads to a lot of shrinkage due to the collapse of the C-S-H structure.
  4. Bound water: This is chemically bound to the hydration product, and can only be removed on ignition. Also called ‘non-evaporable’ water.

2 and 3 are together called ‘gel’ water.

Calculation of the structure of hydrated cement

Theoretically, 0.23 g of bound water is required to completely hydrate 1 g of cement. The remaining water fills up the pores within the structure of the hydrated cement paste (hcp), called the gel pores, as well as the pores external to the hcp, called the capillary pores.

Upon hydration, a volume decrease in the amount of 25.4% of the bound water occurs in the solid hydration product. The characteristic porosity of the hydrated gel is 28%.

Using the above data, some sample calculations are provided below.

Scenario 1: w/c = 0.50; Assume 100% hydration and no drying; Calculate the volume of capillary pores.
Let mass of cement = 100 g. Hence, Vcem = 100/3.15 = 31.8 ml
Mwater = 50 g, therefore Vw = 50 ml.
Vbound-w = 23 ml
Hence, Vsolid-hcp = 31.8 ml + 23 ml – 0.254 x 23 ml = 48.9 ml
Porosity = 28% = 0.28 = Vgel-pores / (48.9 ml + Vgel-pores) —›Vgel-pores = 19.0 ml
Hence total hcp volume = 48.9 + 19.0 = 67.9 ml
Total reactant volume = 31.8 + 50 = 81.8 ml.
Therefore, volume of capillary pores, Vcap-pores = 81.8 – 67.9 = 13.9 ml
Of these, (50-23-19) = 8 ml will be filled with water, and the remaining (5.9 ml) will be empty.

From the above scenario, 23 ml + 19 ml = 42 ml of water is required for complete conversion of 100 g of cement to the hydration product. In other words, a w/c of 0.42 is required. What would happen if the w/c is less than 0.42? Consider the next scenario.

Scenario 2: w/c = 0.30; Cement = 100 g, water = 30 g; Assume that p grams of cement hydrates.
Hence Vsolid-hcp = p/3.15 + 0.23p – 0.254 x 0.23p = 0.489p

Porosity = 0.28 = Vgel-pores / (0.489p + Vgel-pores) (5)
Total water = 30 ml = 0.23p + Vgel-pores (6)

Solving (5) and (6), p = 71.5 g, and Vgel-pores = 13.5 ml
Thus, Vhcp = 0.489 x 71.5 + 13.5 = 48.5 ml
Vunhyd-cem = (100 – 71.5)/3.15 = 9.1 ml
Hence, Vcap-pores = (100/3.15 + 30) – (48.5 + 9.1) = 4.2 ml

That means there are 4.2 ml of empty capillary pores. If this cement paste gets any external moisture (for example, from curing) more cement will hydrate and fill up this space.

Structure of cement hydration products

The structure of C-S-H is best described by the Feldman-Sereda model, shown in Figure 8. It consists of randomly oriented sheets of C-S-H, with water adsorbed on the surface of the sheets (adsorbed water) , as well as in between the layers (interlayer water), and in the spaces inside (capillary water). Such a model implies a very high surface area for the gel. This is indeed found to be true. Using water sorption and N2 sorption measurements, a surface area of 200000 m2/kg is reported (ordinary PC has a fineness in the order of 225 – 325 m2/kg). Small angle X-ray scattering measurements show results in the range of 600000 m2/kg. The corresponding figure for high pressure steam-cured cement paste is 7000 m2/kg, which suggests that hydration at different temperatures leads to different gel structures. The structure of C-S-H is compared to the crystal structure of Jennite and Tobermorite. A combination of the two minerals is supposed to be the closest to C-S-H.

 Figure 8: Feldman-Sereda model for CSH

Calcium hydroxide deposits as hexagonal crystals. These crystals are typically aligned in the long direction inside pores and around aggregate surfaces.

The structure of ettringite consists of tubular columns with channels in between the columns. The imbibing of water in these channels can lead to substantial expansions. Ettringite demonstrates a trigonal structure, while monosulfate is monoclinic.

Figure 9 depicts the relative sizes of pores in concrete. At one end of the scale are entrapped air voids, while on the lower extreme are the interparticle spaces between sheets of CSH.

 Figure 9. Ranges of pore sizes in concrete

Chemistry of special cements

 Expansive cement

 These are cements based on mixtures of Portland cement clinker with expansive compounds. Upon hydration, the typical products that form and cause expansion are ettringite and calcium hydroxide (such as that resulting from free CaO). Many different types of expansive cements are available:

Type K: PC clinker + expansive clinker + gypsum (or gypsum-anhydrite mixture)
The expansive clinker in Type K cement is fired separately. It is composed of a mixture of Alite, Belite, Ferrite, anhydrite and Klein compound, which has the formula C4A3S . In the early stages of hydration, Klein compound reacts faster than C3A to form ettringite. The formation of extra ettringite in the plastic state leads to the initial expansion, which is able to overcome the shrinkage that results from drying.

Type M: PC clinker + Calcium aluminate cement + calcium sulphate (anhydrite)

Type S: High C3A Portland cement with additional gypsum. In general, it is difficult to control the rate of ettringite formation from C3A.

Type O: Portland cement clinker + mixture of Alite, CaO, and anhydrite. In this case, the expansion results from the hydration of the free lime. The CaO is present largely as inclusions within the alite grains and undergoes hydration more slowly as the alite hydrates, resulting in controlled expansive properties.

Supersulphated cement

This cement is a mixture of blast-furnace slag, PC clinker, and calcium sulphate. The amount of blast furnace slag is usually in the range of 80 to 85 % (not less than 75%), while calcium sulphate is added in the amount of 10 – 15%. Overall, the SO3 content of this cement is controlled to be always greater than 4.5%.

This cement requires more water for hydration compared to Portland cement. It is also more susceptible to deterioration during storage due to carbonation. The heat of hydration is lower than PC. When the temperature in service exceeds 50 oC, these cements show a drop in strength, possibly as a result of some changes to the crystal structure of ettringite, which is primarily responsible for the initial strength and stiffening.

The absence of CH in the hydration product and conversion of all aluminous compounds into ettringite during the initial stages makes this cement highly resistant to sulphate attack.

The formation of ettringite is affected by the quantity of lime (CH) available. For a proper reaction, neither too high nor too low of an amount of lime is required. When the amount of lime is too low, carbonation might combine all of the lime available, which is not good. Too much lime (as when there is too much PC in the cement) will interfere with the reaction between slag and calcium sulphate.

Calcium Aluminate cement (or High Alumina cement)

This cement contains 32 – 45% Al2O3, about 15% iron oxides, and 5% SiO2, with the remainder composed of CaO. The primary phase present is Calcium Aluminate, or CA. This cement is produced by sintering a mixture of aluminous (typically bauxite) and calcareous components, and grinding to a fine powder. A complete fusion of all the compounds occurs in the kiln itself, and thus this cement is also called ‘Ciment fondu’ in French.

The types of hydration products that form are dependent on the temperature of the system. When the temperature is less than 10 oC, CAH10 is the hydration product, while between 10 and 27 oC, CAH10 and C2AH8 form. Both these phases, however, are unstable, and a conversion to the stable phase C3AH6 occurs when the temperature exceeds 27 oC. In the long term, gibbsite (AH3) also forms.

The setting time of this cement is similar to PC. The initial strength gain is much faster than PC. For hydration at ambient temperatures, the strength is due to the filling up of pore spaces by the metastable hydration products such as CAH10 and C2AH8. There is a decline in strength when the temperature increases and a conversion to C3AH6 occurs. This conversion can also occur as a result of ageing. The loss in strength due to conversion is a result of an increase in the porosity of the system. The long term strength is due to C3AH6 and AH3. Apart from strength, the durability of the cement is also compromised due to this conversion. CAH10 is inert with respect to sulphates, but C3AH6 can react with SO42- in the presence of lime to form ettringite. The increase in porosity also increases the permeability of the system. The degree of strength loss is dependent on the w/c of the system. At higher w/c, the strength loss is greater.

At extremely high temperatures (such as those found in furnaces and kilns), a ceramic bond can develop between the hydration products and fine aggregate. This lends a very high durability at high temperatures. Thus CA cement is a popular choice for refractory linings.



This site is best viewed in 1024 x 768 resoulution