Mineralogy of binders & the effects of free lime content and cement addition in lime mortars

This document is meant to assist the user in evaluating the composition of various binders and mortars.

In restoration and conservation, it is of the utmost importance that mortars are compatible, chemically and mechanically, with the existing mortars and with the building materials present. This is especially the case in interventions involving the placing of mortars within a structure (re-building, injection, grouting, re-pointing) as potentially damaging chemical or organic components in the binders and the aggregates adopted will cause the deterioration of the structure itself.

The main damaging components are:
– Soluble salts (sulphates are the main ones) > represented mainly by SO3, SO4, CaSO4 (gypsum).
– Aluminates > mainly tricalcium aluminate (C3A)
– The presence of sulphates and aluminates will cause, in a relatively short period, a sulphate attack when these components are exposed to rain water
– Alkalis (K and Na) > their presence, even if generally in small percentages, can cause a reaction with the silica content of the sands (Alkali-Silica reaction) with similar effects to a Sulphate attack.
– Organic content > it must be below 1% of the binder content. Organic matter is subject to deterioration in time and can also promote biological growth.

Mainly because of the above it is by far preferable to adopt pure lime mortars in the restoration or conservation of our built heritage (vernacular or monumental).

The two categories of pure limes

1) Calcium/magnesium (dolomitic) limes – Air limes
Their hardening is the result of re-absorption of CO2 from the air (carbonation), therefore they are classified as Air limes under the current EN/BS 459.1. They are produced from the burning and slaking of calcareous*/dolomitic stone with little or no content of minerals such as silica, alumina and iron oxide.

* a stone is defined as calcareous when it contains more than 50% Calcium carbonate (CaCO3)

The hardening is slow and can be further impeded by the presence of moisture (dampness or rain) which will not allow the surface of the mortar to absorb CO2 due to the formation of a “film” on the surface itself. To obtain a set, in these products, addition of cement and possibly pozzolanic materials are made. These materials such as furnace slag, fly ashes, brick dust, residuals from ceramic production and so on are all grouped under the generic term of Pozzolans and their adoption is nearly always justified by the use of a volcanic ash that the Romans adopted. The ash was from Vesuvius and was called Pozzolan because it was quarried in Pozzuoli, near Naples.

2) Hydraulic limes
So called because they set in contact with water (secondary hardening also takes place in contact with air due to the presence of air lime in their composition). They are produced from the burning and slaking of calcareous stone containing silica, alumina and sometimes ferrite.* The resulting calcium silicates, aluminates and ferrites constitute the hydraulic component.

* As little as 2% silica content in the stone is sufficient to produce hydraulic characteristics. Silica represents 60% of the earth crust and therefore its presence in stones and clays is quite normal. Alumina and iron oxide content is normally much less than silica.

The EN/BS 459.1 does not allow any additions to the pure and natural hydraulic limes (NHL). If cement or pozzolans are added in a quantity of up to 20%, the suffix “Z” has to be added and printed on the bag (NHL-Z). If the addition is over 20% or if various products such as cement, air lime, fillers etc. are blended to produce a binder that sets in contact with water (hydraulic), the product is called HL.

Mineralogical difference between natural hydraulic lime (NHL), high calcium lime (CL/DL) and grey cement

The main components of NHLs are:

1) Hydraulic elements
Predominantly C2S (Belite)
– Some C3S or Alite (derived from possible “high spots” during burning)
– Calcium aluminates (Ca. 2%)
– Ferrites (< 1%)
– Un-combined reactive silica. Portion of SiO2 which has become reactive (amorphous) due to exposure to heat but has not combined with the CaO to form calcium silicates (CS). See Annex 1 (Summary of formation and effects of reactive silica).

2) Non hydraulic elements
Available lime (free lime), Ca(OH)2 or Portlandite.

3) Other Elements
Alkalis (Na,K) between 0.04% and 0.07% for Na2O and 0.10%to 0.25% for K2O.
– Sulphates (SO4) derived from the natural clay content in the stone or from the type of fuel or coal used in the furnaces. As for the Alkalis, the lower the figure the better.
– Unburned residuals. This fraction is inert.

The main components of high calcium limes are:
– Calcium hydroxide – Ca(OH)2 between 70 and 95%
– CaCO3 – limestone residual

The main components of grey cement are:
– Predominantly C3S
– Some C42S
– Aluminates 10%-15%
– Gypsum /CaSO4 : 5% to 10%
– Alkalis (NaK) > 2%
– Fillers

The main differences between NHLs and cement are:

A. Chemical

A.1 – Ratio between C2S (Belite) and C3S (Alite) and corresponding differences in compounds formed during hydration
The hydration of both compounds produces the group CSH or Calcium Silicate Hydrates resulting in similar mechanical performances but C3S will produce 3.2 times more Portlandite [Ca(OH)2] than C2S. See full chemical reaction path in Annex 2.
This Portlandite will be produced as soon as water is added to the mortar and in time, it will crystallise in the pores of the mortar leading to reduced vapour permeability more commonly described as breathability.
The crystallisation of Portlandite will also alter the elasticity moduli of the mortar and stiffens the mortar creating high risk of long term crack formation. The presence in NHLs of amorphous/reactive silica not combined with the CaO is beneficial as it fixes Portlandite, limiting the risk of effluorescence caused by Ca(OH)2 (Portlandite) leaching.

A.2 – Effects of high aluminate content (C3A) and sulphate (Gypsum) addition to cement
Aluminates produce delayed ettringite* by reaction with sulphates which are contained in the binder or in the building fabric and rain water. Sulphates addition in cement is necessary to slow the otherwise near instantaneous set. Otherwise known as “sulphate attack”, this most common problem causes the deterioration of the mortar. The expansion force of the ettringite salt is much higher than the cohesion force of the mortar (70-240MPa versus 4-6MPa), causing cracks, joint delamination with consequent water ingress etc. For the full chemical reaction path see Annex 3.

*Ettringite: calcium hydrate of aluminate trisulphate, also called Candlot salt.

A.3 – Effects of presence of Alkalis
Although the Sodium oxide (Na2O) and Potassium oxide (K2O) or alkali content in cement seems little (>2%), their presence can trigger the “alkali-silica” reaction, attacking the sand component in a mortar.

Alkalis are added to cement as flux materials to reduce the fusion temperature. They stay in the cement and can dissolve silica, causing deterioration of the sand used in the mortar.

A.4 CO2 re-absorption
Cement does not re-absorb any CO2 emitted during production. As a consequence, blended cement/CL binders and mortars have a lower CO2 re-absorption value than NHL binders and mortars.

Table 1: Summary of CO2 emission

Product CO2 Measured in Kg per Ton of Product
Total CO2 emission (from fuel+ decarbonation)
CO2 re-absorbed during Carbonation
Total CO2 not re-absorbed
Hydrated Lime (CL)
St. Astier NHL 5
St. Astier NHL 3.5

The CO2 emission (CO2 not re-absorbed) of high calcium limes and NHLs is quite similar. This is due to the higher total emission in the production of CLs.

B. Physical

B.1 – Rheology, hardening kinetic

The great advantage of an alitic binder (cement based binder containing 3S as its main component) is its fast setting and hardening compared to belitic binders (NHL binder containing mainly C2S or Belite). However, the consequence of achieving high mechanical performance rapidly is higher stress on the support and higher risk of shear. A belitic binder strength gain is more gradual as shown in the following table.

Table 2: Comparison of hardening of Belite and Alite (hardening of pure components expressed in Kg/cm2)

Hardening Comparison 7 Days 28 Days 180 Days 365 Days
Alite C3S
Belite B-C2S

From the table above we see that alite long-term strength is higher than Belite. However, alitic binders (cement based) are much less elastic and breathable than belitic binders (NHLs) as shown in the paragraphs that follow.

B2. – Elasticity moduli

The crystallisation of Portlandite produced in the hydration of C3S (as seen in A1) and by adding CL to cement (1:1:6 mixes etc.) has the effect of stiffening the mortar.
A hard mortar will not accept movements. Construction joints are necessary not only in pure cement mortars but in all blended mortars containing cement.

B3. – Permeability to vapour (breathability)

Permeability is extremely important in restoration and conservation as it diminishes condensation and the consequences of damp. The void structure of cement mortars, especially when badly graded sands are used, is poor. The mortar is quite dense.
Its ability to allow air movement in a structure is much less. Expressed in grams of air x m2 x hour, concrete will have a value of 0.15 whilst NHL 5 and NHL 3.5 mortars made with ISO 679 standard sand at 1:2 ratio reach 0.55 and 0.64 respectively (up to 4 times better).

B4. – Capillarity and water absorption

Cement mortars have a lower capillarity and water absorption than NHL or CL mortars, but their density and correspondent low permeability will promote moisture accumulation within the structure. High capillarity and water absorption are partly compensated by evaporation when a mortar has a good void structure and high permeability to vapour. This is however not the case in cement mortars with their typical poor void structure.

B5. Bonding strength

Cement mortars have a lower capillarity and water absorption than NHL or CL mortars, but their density and corresponding low permeability will promote moisture accumulation within the structure. High capillarity and water absorption are partly compensated by evaporation when a mortar has a good void structure and high vapour permeability. This is however not the case in cement mortars with their typically poor void structure.

The higher bonding strength of cement mortars does not necessarily mean higher durability. Dense mortars can (and do) delaminate if, for any reason such as shrinkage or movement cracks, there is moisture penetration and frost heave. The bonding strength of NHLs is very adequate and the higher permeability and elasticity will result in better durability.
The high bonding strength of cement mortars can also cause damage to the building elements when attempts are made to remove these mortars or when they come off by delamination. The face of a brick, for example, will often remain attached to its cement render. If a brick loses its face, it will absorb much more water and the resulting high suction can cause problems even if the cement render is replaced by a lime render.
Masonry elements built or rendered with cement mortars are rarely recyclable (in the sense of being re-usable) as breakages occur when trying to take the cement mortar off.

Main characteristics and properties of NHL, NHL/CL and OPC/CL mortars related to free lime content (available lime)

FREE LIME CONTENT – %Ca(OH)2 – directly influences a number of the properties of mortars such as: plasticity, fineness, water demand, micro fissures, compressive strength, suction, water absorption, capillarity, vapour permeability, carbonation, lime leaching and bonding strength.

Table 3

Saint-Astier® lime mortars at 1:3 volumetric ratio CL mortars 1:3 volumetric ratio CL + cement blends OPC:CL:sand (binder/sand ratio 1:3)
NHL 5 6.6%

NHL 3.5 8.2%

NHL 2 16.5%
23.1 to 30%
1 : 1 : 6 16.5%

1 : 2 : 9 21.97%

Water lubricates the binder particles, determining plasticity. Workable site mortars have an average consistency with a flow between 180mm and 210mm (vibrating table – EN/BS 1015-3 test is conducted at 190mm). Flow values below or above these figures will produce, respectively, unworkable or unusable, fluid mortars.

CL binders are finer than NHLs (average 10 microns versus 50 microns) so more water is required with CL binders to achieve the necessary plasticity as it has to cover a greater number of particles.

The greater the amount of water, the greater the shrinkage which in turn will cause higher suction and loss of strength. The simple explanation is that water evaporates, creating voids, resulting in shrinkage stress within the mortar. When this stress is superior to the cohesion strength, micro cracks occur. At the surface these are barely visible but they remain within the mortar matrix. CL binders demand more water and its evaporation will increase the shrinkage stress within the mortar. This results in weak mortars that will require protection against frost over a long period (often up to 1 year, depending on the mortar thickness and the exposure zone) as these mortars rarely reach a compressive strength over 2 N.mm2.

The induced porosity described also produces higher suction due to the capillary system of the micro cracks, allowing rain water penetration which not only slows carbonation of the free lime but also favours frost/thaw heave in cold climates. It is the capillary structure that promotes water permeability (absorption). The micro cracks alone represent only a small part of the pore structure of the mortar.

The effect of high free lime content on capillarity and water absorption is quite clear in Table 4 where, for example, an OPC/CL blend has higher capillarity and shrinkage when the free lime content is increased (1:2:9 v 1:1:6) and lower permeability when cement is present due to Portlandite crystallisation within the void structure.

Table 4: Capillarity/shrinkage/vapour permeability Binder:sand ratio 1:3

Capiliarity/shrinkage/vapour permeability Saint-Astier® mortars NHL/CL mortars 50/50/3 CL + cement blends OPC:CL:sand
Capillarity: g.min (@ complete carbonation
4.61 6.30 8.7
NHL 5 NHL 3.5
12.94 13.75
1:1:6 1:2:9
1.08 6.86
Shrinkage mm.m 28 days
0.15 0.25 0.51
0.84 0.89
0.63 0.42
Vapour permeability g.air.m2.h @ complete carbonation
0.52 0.72 0.72
0.63 0.68
0.23 0.25

Pure CL mortars have a higher vapour permeability than NHL and blended mortars which would partly compensate for higher water absorption but they are more susceptible to adverse climatic conditions for a much longer period of time. Their hardening time is prolonged and they need several interventions to prevent high shrinkage.

Carbonation of free lime is directly related to the mortar’s exposure to air. Simply explained, limestone (CaCO3) releases CO2 during burning and becomes CaO. To harden*, it needs to re-absorb CO2 from the air, returning therefore to CaCO3. This process is called carbonation. Factors impeding carbonation are surface damp (formation of a surface patina impeding contact with air), surface water (rain), non-breathable surface coatings and finishes. Mortars placed within structures (consolidation/grouting mortars) or building/re-pointing thick walls will also have poor or much slower carbonation in the areas not near to the surface of the joints. The presence of un-carbonated free lime causes lime leaching if water is also present because un-carbonated free lime is partly soluble in water at a ratio of 1.6 g/litre (less in presence of other salts).

* Hardening is not to be confused with setting. It takes much longer and is related to carbonation of free lime and, in hydraulic binders, full development of belitic and alitic reactions.

The bonding strength of mortars, so important in rendering and plastering work, is also related to the free lime content in the binder.

Table 5

Saint-Astier® lime mortars binder/sand ratio 1:3 NHL/CL mortars 50/50/3 binder/sand ratio 1:3 CL + cement blends OPC:CL:sand binder/sand ratio 1:3
NHL 5 0.51

NHL 3.5 0.46

NHL 2 0.36
NHL5/CL 0.28

NHL 3.5/CL 0.22
1 : 1 : 6 0.7

1 : 2 : 9 0.5

The average bonding strength of a pure CL 90 mortar will be much lower than the 50/50 NHL-CL mortars shown in Table 5. It might reach the minimum requirement of 0.3 N.mm2 at full carbonation, but this would expose the work to potential weather damage for too long a period if one considers that the carbonation of pure CL is about 5-10mm per year.

Although the presence of free lime helps the workability of a mortar, the higher the amount of free lime, the higher the water demand, capillarity, water absorption shrinkage, setting and hardening time. The bonding strength and frost resistance are lower.

The higher the free lime content, the more the application season and protection of these mortars must be considered, especially in high rain exposure zones and in cold climates.

The addition of pozzolanic materials will enhance the setting properties of mortars with high free lime content but not much is known about the long-term effect of pozzolanic additions on the other properties and the overall durability of these mortars.

If, as is the case of a number of widely used mortars, cement is added to speed up setting, there will be detrimental effects due to the composition of cement and the properties of cementitious mortars as shown earlier in this document.

In conservation and restoration it is essential that new materials are compatible with the existing ones. In many situations weak mortars or mortars with a high free lime content are required. When looking at the suitability of these mortars the following points should be considered:
– Site exposure
– Application season
– Protection and curing
– Condition of the building structure (good or friable etc.) and its detailing
– Wall dampness and its causes
– Presence of salts
– Choice of aggregates
– Void structure, vapour permeability and capillarity requirements of the new mortars.
– Thickness and depth of the joints (carbonation time of building – pointing mortars) or size and structure of voids (injection and grouting mortars).
– Thickness of the render (thin renders with mortars with high free lime content will carbonate quicker than thick renders).
– Workmanship and ability of the contractor to comply with the specifications issued.

When dealing with ancient structures further consideration has to be given to:
– The function of the new mortar in relation to the state of the structure. This is particularly relevant to the conservation of ruins where weak mortars might have been used to construct the original structure but the objective now is to preserve what is left. A chemically compatible stronger mortar could be the correct choice, especially in very exposed sites.
– The potential loss of binder in the existing mortar.
– The correct individuation of the presence of soluble silica (hydraulic element) in the existing mortar.
– The durability requirement in relation to possible future interventions or maintenance (from “sacrificial” to “perennial”).

The responsibility of the specifier is to understand the intervention needed and choose the most suitable mortar/s for the work to be done. The responsibility of the supplier is to provide correct and in-depth information to assist the specifier in making the correct choice. The responsibility of the contractor is to provide the necessary skills and follow the correct working practice.

Potential failures are possible if the composition of the binder/s to be used is not known or understood, if additions (cement, pozzolans, additives) are made without knowing their long-term effects, if superficial knowledge is mistaken for professional knowledge and if price considerations prevail on technical ones.

Ugo Spano/Laurent Tedeschi
September 2006


Annex 1: Summary of formation and effects of reactive silica

All natural limes are produced in industries by calcination of calcareous rocks at a temperature ranging from 900\B0C to 1300\B0C. Below these temperatures, carbonation (1) yield is not sufficient. At higher temperatures, lime crystals will combine back and the resulting quick lime will be hard to slake. For natural limes with hydraulic properties, over-burned CaO is formed which explains how slaking NHL lime is slower than slaking CLs. The slaking reaction of NHL passes through a maturing phase as described by M LE CHATELIER. In this phase all CXS are preserved.

(1)    Ca CO3 + Δ (Energy) → CaO + CO2

If calcium oxide or quick lime is in contact with silica, combination will result. This reaction is exothermic

(2)    2CaO + Si 02 → (CaO)2 Si O2

In other words, this reaction gives energy as opposed to a decarbonation reaction (1) which is endothermic (needs to take energy from the system to occur). It is necessary to heat limestone to produce lime but as soon as this lime is produced, it combines exothermically with the hydraulic elements present.

Limes are generally produced in vertical kilns so material must stay in solid form through the whole length of the kiln unlike cement production where horizontal kilns are used and material is transformed through a liquid state (fusion).

Silica is present in limestone as small inclusion particles (>1 mm) and has only a superficial contact with calcium oxide. Combination reactions are only produced on the surface of silica particles. The internal silica particles do not combine but have been heated and the surface combination reaction is exothermic (2).

Original silica particles in the quartz crystalline state (sand) change to an amorphous state (pozzolan, basalt, pumice) by the effect of heat. When these are crushed through grinding, amorphous silica becomes available. This amorphous silica is called “pozzolanic”. A very basic (pH ~ 12.5) mixture that converts amorphous silica to liquid, as indicated by Boyton, is obtained when lime is mixed with water to prepare a mortar,

(3)  ≡Si – O – Si≡ + 3OH- → [SiO(OH)3]- .

It combines with free lime to produce:

(4)   Y [SiO(OH)3]- +X Ca2+ + (Z-X-Y) H2O + 2(X-Y) OH-  →  CxSyHZ

The “types” of CHS (x, y and z values) obtained after reaction are related to the concentration of the present species but these crystals do not correspond to CSH obtained by belite (C2S) crystallisation. Theoretically these “types” can be identified but in practice no exhaustive studies have been conducted.

This pozzolanic gain is active or useful since the increase of mechanical performance can only be efficient if there is some hydraulicity in the system which is the case for natural hydraulic limes. For calcium limes, this hydraulicity is introduced by the addition of cement or from endogenous activators in crushed pozzolans [Na2O, K2O (alkalis) with SiO2 in pozzolans].

Schema 1

In this table, a mixture of pozzolan is represented with no sand since it would be dissolved by alkalis.

To obtain measurable performance increase, CxSyHz must have seeds of CSH to grow. This is the crystalline germination step.

In conclusion: there will be no pozzolanic activity if there was no initial hydraulicity. Addition of amorphous silica to calcium lime will not provide measurable mechanical performance increases whereas additions to NHL 2 can result in a very high compressive strength increase.

In an NHL mortar, the influence of amorphous silica not initially combined to form calcium silicates but made available through the crushing of some oversized slaked hydraulic lime granules is positive since it fixes the Portlandite [Ca(OH)2] formed in the hydration of C2S and limits the risk of surface efflorescence caused by leaching of Ca(OH)2.

Annex 2: Difference in C2S v C3S hydration and Portlandite formation

During mortar setting, CxS (C2S and C3S) will hydrate to produce « C-S-H » compounds according to the following equations (1) and (2)

Alite hydratation (C3S) :

(1)   2C3S + 10.6 H → C3.4-S2-H8 + 2.6 CH

2(CaO)3 SiO2 + 10.6 H2O → (CaO)3.41 (SiO2)2, (H2O)8 + 2.6Ca(OH)2

MC3S = 228 g.mol-1 MCH = 74 g.mol-1

1 % in mass of C3S will produce          74*2.6/228*2 = 0.42% mass

Belite Hydratation (C2S) :

(2)   2C2S + 8.6 H C3.4-S2-H8 + 0.6 CH

2 (CaO)2 SiO2 + 8. 6 H2O (CaO)3.41 (SiO2)2, (H2O)8 + 0.6 Ca(OH)2

MC2S = 172 g.mol-1 MCH = 74 g.mol-1

1 % in mass of C2S will produce      74*0.6/172*2 = 0.13% mass

Comparison of equation (1) and (2) leads to the following observations:

C3S and C2S will hydrate to form the same CSH, resulting in similar mechanical performances but C3S will produce 3,2 times more Portlandite than C2S, increasing the risk of lime leaching.

Annex 3: Aluminates and sulphates > sulphate attack

Aluminates can produce “delayed” ettringite by reacting with sulphates that are contained in old masonry work (gypsum roughcast, gypsum renders on walls, capillary effect from sulphate water) or coming from urban pollution. Ettringite is a calcium hydrate of aluminate trisulphate with the following formulation:

C6AS3M32 [(CaO)6AI2 O3(SO3)3(H20)32] also called Candlot salt

Sulphates are added to cement to control its set time since the aluminates contained in cement have a very rapid set (measured in minutes) which causes a false set.
These sulphates are added in the form of gypsum that is “captured” by the aluminates at the beginning of hydration to form “premature” ettringite. These reactions are harmless to the final quality of the mortar since its initial set has not yet started.

The following reactions are taking place (1)-(2) and (3)

(1) C3A + 3CSH2 + 26H → C6 HS3H32
Gypsum has been “consumed” by reaction (1) giving: 

(2) 2C3A + C6ASH32 + 4H→ 3C4ASH12

Premature ettringite reacts to give:
(3) C3A + CH + 12 H → C4 A H13

These reactions are normal in a cement based component but if sulphates are added to the mortar after it has set, the following reactions occur:

These reactions are normal in a cement based component but if sulphates are added to the mortar after it has set, the following reactions occur:

(4) CH + SH →  CSH

with C4ASH12 from equation (2) we obtain:
(5) C4 ASH12 + 2CSH + 1 6H → C6 AS3H32

Crystal volume almost doubled 1240 → 2144

With C4 AH13 from equation (3) we get:

(6) C4AH13 + 2CSH + 14H → C6AS3 H32 + CH

Again, crystal volume almost doubles 1148 →  2144

These expansion reactions take place when the mortar has already set and is no longer flexible.

The expansion caused by salt crystallization results in pressure that can be as high as 70 to 240 MPa according to some authors. This stress is considerably higher than the cohesion strength of a mortar, which is in the order of 4 to 6 MPa.

Annex 4: CO2 emission of various binders

Material - Energy Requirements Fuel (coal) Kg/Ton Thermie (Fr) per Ton * Kilowatts/hour Per ton
Hydrated CL 90%
NHL 2 with 75% CL
NHL2 with 60% CL

* 1 Thermie (Fr) = 4.1855 MJ or 0.03967 Therm

Material CO2 per Kg/Tonne of Binder
Emission during burning
Emission from de-carbonation
Total emitted during production
Not Re-absorbed
Hydrated CL 90%
NHL 2 with 75% CL
NHL2 with 60% CL
MORTARS 1:2 binder/sand ratio Weight of binders/kg per ton of mortar Total CO2 emission (not re-absorbed) kg per ton of mortar MORTARS 1:3 binder/sand ratio Weight of binders/kg per ton of mortar Total CO2 emission (not re-absorbed) kg per ton of mortar
Cement/CL 1:1:4
Cement/CL 1:1:6
NHL 2 with75%CL
NHL 2 with75%CL
Saint-Astier® NHL 2 with 60% CL
Saint-Astier® NHL 2 with 60% CL
Saint-Astier® NHL 3.5
NHL 3.5
Saint-Astier® NHL5