Materials and Structures, 1994, 27, 3 9 3 - 3 9 8

Modelling lime mortar carbonation

K. VAN B A L E N , D. VAN G E M E R T
Catholic University of Leuven, Department of Civil Engineering, De Croylaan 2, B-3001 Leuven (Heverlee), Belgium
The carbonation process in lime mortar is influenced by the diffusion of carbon dioxide into
the mortar pore system, by the kinetics of the lime carbonation reaction, and by the drying
and wetting process in the mortar. All these phenomena depend on the presence of water in
the mortar. Water is present in liquid form as well as vapour. Physical models and their
mathematical basis are given for the factors influencing carbonation. The mathematical
description of the carbonation process in lime mortar completes the information necessary to
model the simulation of carbonation in conjunction with the drying process. Simulations at
different carbon dioxide concentrations showed a good correlation between experiment and
mathematical prediction

1. I N T R O D U C T I O N
Although lime mortar has been used for centuries, current
knowledge is insufficient for an evaluation of the
mechanical behaviour of historical masonry. The slow
carbonation of lime m o r t a r causes the mechanical
properties of lime mortar and of the masonry built with
it to evolve with time. Recent research [1] has reported
on this carbonation process and the changes in the
mechanical and physical properties of lime mortar which
are caused by carbonation.
The rate of carbonation is to a great extent influenced
by the presence of moisture in the mortar. Therefore, the
carbonation process should be divided into a carbon
diffusion process followed by a chemical reaction in which
calcium carbonate crystals are formed. The influence of
moisture on carbon diffusion parameters in mortar and
the influence of moisture on the chemical reaction were
studied. A new carbonation model was set up in which
the progress of carbonation takes the drying of the lime
mortar into consideration. The deposition of calcium
carbonate crystals during carbonation changes the pore
structure. Current research shows that this structural
change influences the hygroscopic properties of lime
mortar.
2. C A R B O N A T I O N R E A C T I O N

retarded when the material is saturated with water. The

optimum water content for carbonation is the water
content that corresponds to maximum adsorption on the
surface of the pores before capillary condensation. The
water content of the material therefore plays an important
role in the diffusion of carbon dioxide in the lime mortar.
3. F A C T O R S I N F L U E N C I N G

CARBONATION

The approach followed distinguishes between mass

transport of carbon dioxide, mass transport of water
vapour and the carbonation reaction.
3.1 Factors related to diffusion of carbon dioxide
3.1.1 Time
As for many diffusion problems, the progress of carbonation can be expressed in a simple form:
X = kx/t

or

x = e + kx//t

(3)

where x is the carbonation depth; k is a factor; t is the

time; and e is a constant. The k factor is often determined
in an experimental way based on one of the above
expressions (3). This k factor does not correspond to a
property of the material, although many researchers are
looking for its relationship to material properties and
using its value to predict carbonation speed [3].

The carbonation reaction proceeds in two stages.

3.1.2 Porosity and diffusivity of the material

1. Carbon dioxide is dissolved in water

Both material properties are interconnected, as the

diffusivity depends on the pore structure of the material.
The pore structure of the material also affects the
hygroscopic behaviour, and as such influences the effect
of water content on the diffusion of carbon dioxide in
the material.

CO 2 + H 2 0 ~ H2CO 3

(1)

2. Lime reacts with the carbonic acid

Ca(OH)2 + H 2 C O 3 =~ CaCO3 + 2 H 2 0

(2)

This is confirmed by the slow carbonation of concrete

at low relative humidities [2]. Diffusion of carbon dioxide
in water is much slower than in the air. The binary
diffusion coefficient in the air is 1390 x 10 - s m 2 s - i ; in
water it is 104-105 times less [3]. Carbonation is thus
0025-5432/94 9 RILEM

3.1.3 Construction method and cracks

The carbonation speed of site-mixed concrete is faster
than that of prefabricated concrete which, in turn, is

394

V a n Balen and Van G e m e r t

faster than that of prestressed concrete [3]. In the case

of masonry, carbon dioxide can also penetrate into the
material through the openings between the mortar and
the joints and also through the bricks, as was noted in
a British study [4]. O u r research is confined to onedimensional diffusion in a material assumed to be
homogeneous.
3.1.4 Geometricalshape
In practice, carbonation is often two- or three-dimensional.
The carbonation front therefore is not always a plane
surface. A greater external surface in relation to the inner
mass speeds up carbonation as well as drying in the case
of two- or three-dimensional diffusion.

3.2 Factors related to the carbonation reaction

This concerns mainly the lime and the reactivity. The
progress of the carbonation front depends on the amount
of lime, which has to react with the carbon dioxide
diffusing in the material. A minimum amount of water
in which the carbon dioxide has to be dissolved is needed
for the reaction. The solubility of carbon dioxide and
lime decreases with increasing temperature, whereas
chemical reactions in general are faster with increasing
temperature. Therefore, the optimum carbonation speed
is found at about 20~ [5]. The solubility of lime depends
on the particle size. The dissolution of lime, which we
suppose to be all at the pore surface, is faster than the
slow diffusion rate of carbon dioxide. We therefore can
count on a maximum Ca(OH)z content in the water on
the pore surfaces as long as not all the calcium hydroxide
is carbonated.
The dissolution of carbon dioxide in water depends on
the transition coefficient for mass transport between
water and air. This coefficient is unknown but, under
equilibrium conditions, a relation exists between the
carbon dioxide concentration in water and the partial
pressure in air:

3.3 Factors related to drying and wetting

The climatic factors considered were relative humidity,
wind speed, (rain)water and temperature. The climate
affects the carbonation since it influences the water
content of the material by wetting and drying. The drying
process is influenced by the wind speed since it influences
the mass transfer coefficient, by the temperature and
relative humidity of the ambient air. The optimum drying
of fresh mortar, after which carbonation can start, is
favoured by low relative humidity, strong wind velocity
and high temperature.

4. M A T H E M A T I C A L D E S C R I P T I O N
OF CARBONATION
Carbon dioxide diffusion in the air is described by Fick's
first law. To describe the diffusion in porous materials
the diffusion coefficient is replaced by an effective
diffusion coefficient in the material [3]. Carbon dioxide
diffusion is possible only in the open pores. The pore
volume is not constant and changes with the water
content. Carbonation itself modifies the pore volume. It
has been shown that the total pore volume decreases with
carbonation [8]. Our own experiments have shown that
in lime mortar the relative amount of narrow pores with
a diameter between 0.0063 gm and 0.1 pm diminishes
with carbonation. The decrease of porosity and the
change of pore size distribution is very small, so that we
may assume that the porosity ~ is independent of the
degree of carbonation and is dependent only on the water
content. The relation between porosity and water content
is deduced from the hygroscopic curves (isotherms). The
ratio of the hygroscopic water content (WH) to the
theoretical water content at 100% RH may be applied
to the total porosity. This method allows taking into
account the enclosed air which does not increase the
water content but leads to a decrease in the apparent
porosity.

~(W)=~max(1

[CO2]w~-.KHPk
[C02]w
[CO2]
[CO2]w
Kn

Pk

KHRT[CO2]J

carbon dioxide concentration in the air (mol

or tool m - 3);
carbon dioxide concentration in water (tool
or mol m-3);
Henry constant (10 -15 m o l l -1 atm -1
mol m - i N - 1);
partial pressure of carbon dioxide (atm
Pa).

WH )

(5)

w~

(4)
1-1
l- 1

where 0(w) is the porosity at water content w (m 3 m - 3),

0m,x is the absolute porosity of the material (m 3 m-3),
wa is the water content at RH (kg m-3), and w100% is
the water content at 100% RH (kg m - 3).

or
or

The result of carbonation is the formation of calcium

carbonate crystals, mostly calcite. Carbonation with
high carbon dioxide concentration leads to the formation
of aragonite crystals [6]. This has been noted on
samples carbonated in a 100% CO2 atmosphere. The
presence of SO2 can slow down the reaction considerably
[7].

4.1 Carbonation reaction

In a volume dxdydz, part of the carbon dioxide is consumed by the reaction. For every mole of Ca(OH) 2 one
mole of carbon dioxide is taken up and one mole of water
is produced. Formula 6 expresses the conservation of
mass, neglecting the molecular adsorption and the change
of the carbon dioxide concentration in the air:

t~gkx ~gky+ 6gkz 6x @

6z

- - +

R(w,c)-~--

~C
6t

(6)

Materials and Structures

395

with
0 = D~ Vc
#
div

dr0

Vc + R(w,c) + tP dt

where g is the mass transfer of carbon dioxide, c is the

carbon dioxide concentration, # is the diffusion resistance
number, Dk~ is the diffusion coefficient of CO2 in air
(1390 x 10 - s m 2 s - l ) , and R(w, c) is the reaction speed
per unit volume (mol s - 1 m - 3).
This reaction term R(w,c) expresses that, in an
elementary volume dxdydz, R moles per unit of time
and unit of volume of the gas are consumed by the
reaction. The consumed carbon dioxide molecules cannot
participate any more in changing the concentration of
the gas in the air of the pore.
We found that the reaction term R can be expressed as

R = kT[CO2]ct(RV)fl([Ca(OH)21)

(8)

where k r = k2~
k 2 ~ 0.7062 (s -1) parameter at
20~ f ( T ) is the temperature relation; [CO21 is the
carbon dioxide concentration in the air of the pore
( m o l m - 3 ) , fl([Ca(OH)2]) is the ratio of pore surface
taken over by lime to the total pore surface [Ca(OH)2]/
[Ca(OH)2]o, and ~(RV) is the influence of R H on
reaction term (see later).
This is based on the assumption that calcium hydroxide
will be dissolved in the pore water as long as calcium
hydroxide is available at the pore surface, as long as not
all Ca(OH)2 has carbonated to CaCO3. During all that
time the assumed concentration of calcium hydroxide in
that water will remain the maximum concentration that
can be dissolved in water: 23 mol m - 3 at 20~ According to Pinsent [7, p. 1512] the second order reaction
constant is 8.4 m 3 m o l - ~ s - 1 which allows us, estimating
the thickness of the water layer at the capillary pores at
5 molecule layers (1.3 10 -9 m), to calculate the value
ofk 2~ = 0.7062 (s- I) at 2 0 ~
= 1 for the isothermic
carbonation condition assumed at 20~ The autocatalytic
character of the reaction between calcium hydroxide and
carbon dioxide [51 justifies this expression. In [1]
expressions are given for the different parameters of
formula 8.
Water transport at drying can be split into two phases:
capillary water transport followed by water vapour
diffusion. During the first drying phase (see Section 4.2)
carbonation is assumed to be negligible as the pore
structure is not sufficiently open to allow carbon dioxide
to diffuse. This drying phase can be calculated independently of the carbonation, as almost no water is
produced by carbonation. The period of the first drying
phase can be calculated assuming one-dimensional
drying at one surface [91.
When a fully saturated fresh mortar is exposed to
carbon dioxide, we can assume that the bulk of the
carbonation in the material will start after the first drying
phase. The outer surface of the material can start to

carbonate earlier as the transfer of carbon dioxide to the

surface is hindered only by the transfer coefficient, which
is more or less independent of the water content of the
material. Carbonation at the surface is stopped very
quickly and the carbonation will only continue after a
time tkn when the first drying phase has passed and the
second drying phase starts. This approach neglects
carbon dioxide diffusion when the water content in the
material is between the capillary water content and the
critical water content. When the critical water content is
reached, capillary water transport to the surface is stopped.
Then the water front in the material withdraws due to
diffusion of water vapour. Before the critical water content
is reached, the large pores can yet be free of water and
therefore, theoretically, diffusion of carbon dioxide should
be possible. As the first drying phase, compared with the
second one, is very fast, the amount of lime that can carbonate in the period that the water content is between WCap
and W~ris very small. The influence on the accuracy of the
model will thus be very small. Further research is continuing to determine the carbon dioxide diffusion resistance
as a function of the water content of the lime mortar.
4.2 Water vapour transport
Drying is modelled in two stages: drying at the surface
and capillary water transport to the surface (first drying
phase), followed by the diffusion of water to the surface
(second drying phase).
The water vapour transport in the second drying phase
is expressed as a vapour pressure potential, using Fick's
law:
6a
gw, = - - - - Vp = --6wv Vp
(9)
flwv

where 6a is the diffusion coefficient of water vapour in

air, 6wv is the effective diffusion coefficient of water vapour
in the material, and #wv is the diffusion resistance number
for water vapour through the material. The diffusion
resistance number (#wv) is a function of the relative
humidity or water vapour pressure (p). Due to the
shortening of the diffusion path, when the material is filled
with water, # decreases with increasing water content to
become about 1 for w = wcr (critical water content at the
end of the first drying phase). If the water content still
increases, the water mass transfer due to vapour diffusion
becomes negligible compared with the mass transfer due
to capillary water transport.
The number #wv as a function of the R H (r can be
approximated with the following empirical expression:
#wv -

(10)

For lime mortar n = 12, as the hygroscopic curve is very

steep; for cement mortar, n = 2 is a good approximation.
F r o m diffusion tests we know that p = 12 for ~b = 75~o;
/~ = 1 at w , (the critical water content wet for the fresh
lime mortar is 53.4 kg m - 3). F r o m these the values of a

396

Van Balen and Van G e m e r t

13

11

1Q
9
8

1.1

0.9

0 0

JL
0.8
~
o

G"

0.7
0.6

0.5

0,4

6
5

c
~

..3

<

0.2

4
3

0.1

2
1

I
2O

]
I
I
40
60
Re{otive humidity (%)

I
80

IO0

I
40
6'0
Relative humidity (~)

20
[] measured values

Fig. 1 Evolution of #~,, in lime mortar as a function of RH.

--

I
80

100

oppfirnot]ve relation

Fig. 3 Ratio of the available pore surface to the total surface,

taking into account the influence of molecular adsorbed
water and capillary condensation.

and b can be determined, a = 0.0833, and b = 0.9167.

/Z.v

(11)

0.0833+ 0.9167(10~) t2

For lime mortar, this leads to the curve in Fig. 1.

For carbonation a diffusion model with a variable
diffusion coefficient and variable capacity term is needed.
Therefore an approximating expression for the hygroscopic water content w = f ( p ) has been fitted on
experimental results which allows the calculation of f '(p).
Also #w~ depends on the water vapour pressure. The
solution procedure for the differential equations will have
to take this into account.
The relation between the carbon dioxide diffusion
coefficient and RH can be approximated by a curve with
the inverse form of the hygroscopic curve. It reaches zero
between the capillary and the critical water content, at
100% RH. At that water content only isolated water
islands are present in the material, and no diffusion of
carbon dioxide is possible. The diffusion coefficient at
low R H is more or less constant, and can be derived from

~"

1.9
1.8
"L7

1.61.5~.3-

,2
Zone of value of D K
o

0.9-

"Y 0.8
.m
u 0.7
u
=
.s
~

0.5
0.4
0.,3

c~

0.2
0.1
0

50
BO 87
O I 76184 190

94

96

20

~.H.

;0

60

Water content ( k g / m 3)

Fig. 2 Evolution of carbon dioxide diffusion coefficient as a

function of water content in the material.

the diffusion resistance number for water vapour (/~) of

the material. This diffusion resistance number can be
applied to the carbon dioxide diffusion coefficient in the
air (Dk).
The diffusion coefficient of lime m o r t a r can be
expressed as
O~ -

1
(12)

The lime mortars tested gave # = 12 at 75% RH, therefore

Dk = 1.9 x 10 -7 m 2 s -1. At 100% RH, Dk is very small,
and we consider then D~ 1/100 of 1.9 x 10-7 m 2 s - t . It
is shown in Fig. 2. The influence of relative humidity on
the reaction term is calculated in the factor c~(RH)
(0 < e ( R H ) < 1), which expresses which part of the
surface is free for the reaction [I]. The result is shown
in Fig. 3 for the lime mortar samples tested.
5. M A T H E M A T I C A L M O D E L O F L I M E
MORTAR CARBONATION
Carbonation can thus be expressed in the form of a
differential equation with a 'sink' term. The composition
and the structure of the material influence the parameters
in the reaction term and the diffusion resistance number.
For the simulation of the carbonation of lime mortar,
not taking into account the effect of humidity and
assuming that the reaction is much faster than the
diffusion, an analytical expression with a new definition
of the 'sink' term is proposed. This sink term is based
on the concentrations of lime and of carbon dioxide. The
equations for the description of water vapour transport
complete the information needed to construct a model
that allows the simulation of carbonation in conjunction
with the drying process.
The mathematical formulations expressing carbonation
and taking into account the effect of water content of the
material are used to construct a computer simulation
model. The nonlinear coupled differential equations are

Materials and Structures

397
diffusion controlled phase of carbonation starts when the
lime mortar has dried enough. This leads to a curve with
two peaks: a sharp one after a short time and a rounded
one after a longer period, as found in the experiments of
Henkel [10]. The first sharp peak found in the mathematical model is not as clear in most experimentally
measured curves expressing the carbon dioxide uptake
due to carbonation. This is due to the measuring
procedures which are not able to follow the rapid carbon
dioxide uptake at the very early stage after the preparation
of the lime mortar.

~'G 0.008

0.006-

E 0.005-

-8 0.004x
o

ooo2/
:~ 0.003

o.ool
o
co~-o.ool
9
1.0
112
~'
]3me (1000 s)

,o,.2,o'.41o~6Lo'8'

, 1,4 ,

116

118

2.0

Fig. 4 Carbon dioxide intake of lime mortar carbonating in

100% CO 2 and 60% RH environment.
solved using a finite difference procedure incorporated
in a program in Turbo Pascal. This allows us to simulate
the one-dimensional carbonation process of a lime
mortar. It also allows us to study the effect of drying on
the progress of the carbonation front in lime mortar.
With appropriate adjustments, the same model could be
used to simulate the carbonation of concrete.
As an example, the carbonation of freshly mixed lime
mortar is simulated with the mathematical model in
Figs 4 and 5, which give the evolution of carbon dioxide
intake due to carbonation when drying occurs in
conditions of 60% RH. The carbon dioxide intake of
freshly made lime mortars as studied by Henkel [10] has
been simulated. A precise comparison of the simulation
with the data given by Henkel is difficult as in Henkel's
work the carbon dioxide diffusion was not limited to a
one-dimensional diffusion as it is in the model. Therefore
the amount of carbon dioxide fixed in the first phase is
very important.
Figs 4 and 5 show the calculated carbon dioxide flux
at the surface of the lime mortar in environments with,
respectively, 100% and 20% carbon dioxide concentrations. The typical trend of the carbonation intake is
found. It is shown that the carbonation of the lime m o r t a r
outer surface occurs very quickly and that the second
~'G 0.0028

Carbonation can be expressed by means of a differential

equation with a 'sink' term. The composition and the
structure of the material influence the parameters in the
reaction term and the diffusion resistance number.
For the simulation of the carbonation of lime mortar,
not taking into account the effect of humidity and
assuming that the reaction is much faster than the
diffusion, an analytical expression with a new definition
of the 'sink' term is proposed. This sink term is based
on the concentrations of lime and of carbon dioxide.
The equations for the description of water vapour
transport complete the information needed to construct a
model that allows the simulation of carbonation in
conjunction with the drying process.
The mathematical formulations expressing the carbonation and taking into account the effect of water content
in the material were used to produce a simulation model.
The non-linear coupled differential equations are solved
by a finite difference procedure.
The model allows us to study the effect of drying on
the progress of carbonation in the lime mortar. With the
appropriate adjustments the same model can be used to
simulate the carbonation of concrete. This model thus
contributes to a better understanding of the carbonation
process of lime m o r t a r s and also of cementitious
materials.

REFERENCES

0.0026
.~ 0.0024
0.0022
0.0020
c 0.0018

0.0016
._~ 0.0014
~z 0.0012
.Q

6. C O N C L U S I O N S

o.oolo

0.0008

2 0.0006

o
.~ 0.0004
0.0002
Q.
cn
0
0

2i

4~

6~

~
,
8~ ~ 10
1~2
]]me (1000 s)

'

'
14

1~6

'
18

'

20

Fig. 5 Carbon dioxide intake of lime mortar carbonating in a

20% carbon dioxide concentration and a 60% RH environment.

1. Van Balen, K., 'Karbonatatie van kalkmortel en haar

invloed op historische strukturen (Lime mortar carbonation and its influence on ancient structures)', Ph.D.
Dissertation, K.U. Leuven, Leuven 1991.
2. 'CUR-Report 73, Carbonatatie lichtbeton. Literatuurstudie
(Carbonation of lightweight concrete. Literature)' (CURInstitute, Gouda, The Netherlands, 1975) pp. 9-51.
3. Richardson, M., 'Carbonation of reinforced concrete. Its
causes and management' (Dublin, 1988).
4. Hunt, J. H., 'The loss of alkalinity of brickwork mortars.
Measurements taken on buildings at the BRS' (Building
Research Establishment Note, N 149/80, Garston, UK,
1980).
5. Grandet, J., 'Contribution ~ r6tude de la prise et de la
carbonatation des mortiers au contact des mat6riaux

398
poreux', Ph.D. Dissertation, Univ. Paul Sabatier Toulouse,
CNRS, 1975,
6. Roques, H. and Girou, A., 'Kinetics of the formation
conditions of carbonate tartars', Water Res. 8 (1974)
907-920.
7. Sada, E., Kumazawa, H., and Butt, M. A., 'Simultaneous
absorption with reaction in a slurry containing fine
particles (two gases)', Chem. Enong Sci. 32 (1977)
1499-1503.

RESUME
Mod~iisation de la carbonatation du mortier ~i la chaux
On donne un aperqu gdnbral des diffdrents facteurs qui
influencent la carbonatation du mortier d la chaux. On
dkmontre que le contenu d'eau dans le matkriau influence
la carbonatation aussi bien par un changement de la
diffusivitk d travers le matdriau que par un changement de

Van Balen and Van Gemert

8. Kroone, B. and Crook, D. N., 'Studies of pore size
distribution in mortars', Mag. Concr. Res. 13(39) (1961)
127-132.
9. Hens, H., 'Bouwfysica 1. Warmte en vocht: Teoretische
grondslagen (Building Physics 1. Heat and humidity:
fundamentals)', Leuven (1981).
10. Henkel, W., 'Untersuchungen fiber die Kohlens/iureaufname von M6rteln', Zement-Kalk-Gips 11(8) (1955)
386-393.

la rbaction entre le gaz carbonique et la chaux. Un modkle

mathbmatique a ~tk dkveloppb qui permet la simulation du
processus de la carbonatation en tenant compte de
lYvolution de sbchage du matbriau dans son environnement.
Des simulations avec des concentrations de gaz carbonique
diffkrentes montrent une bonne correlation entre les
expkriences et les prbdictions mathbmatiques.