Progress in Organic Coatings 48 (2003) 259270

Criteria and methodology for the evaluation of conservation interventions

on treated porous stone susceptible to salt decay
A. Moropoulou , N. Kouloumbi, G. Haralampopoulos, A. Konstanti, P. Michailidis
Department of Materials Science and Engineering, Faculty of Chemical Engineering, National Technical University of Athens,
Iroon Polytechniou 9, 15780 Zografou, Athens, Greece

Abstract
Susceptibility of stone to salt decay is directly related to its microstructural and mechanical characteristics. In the present work, a
porous stone from a quarry in Cyprus was examined. Samples of this stone were consolidated and protected with certain conservation
materials. The modification of microstructural characteristics of the stone was evaluated by mercury intrusion porosimetry, while its
mechanical characteristics were measured by compressive and bending strength tests. From these data, the susceptibility of Cyprus stone
to salt (NaCl) decay was estimated. The values of crystallization pressure of NaCl were calculated and compared to the experimentally
measured compressive and tensile strength of the stone, in order to examine the probability of stone disruption due to salt crystallization.
The development of the phenomenon after coarse pores filling with crystals, i.e. crystallization of salts in capillaries or mechanical failure
of the stone, was determined by thermodynamic analysis. Additionally, artificial weathering tests of marine salt spray were performed for
the evaluation of stones durability after the conservation. The treated stone proved to be resistant to salt decay, as it presented ameliorated
microstructural and mechanical characteristics, concerning decay due to soluble salt crystallization. The most possible thermodynamic
scenario was small pores repletion with crystals, avoiding mechanical failure of the stone.
2003 Elsevier B.V. All rights reserved.
Keywords: Salt decay; Aging tests; Microstructural stone characteristics; Mechanical stone characteristics; Thermodynamics of crystallization; Conservation

1. Introduction
Susceptibility of stone to certain decay forms (frost, soluble salts) is highly dependent on its microstructural and
mechanical characteristics [1]. Soluble salts are considered
to be the key deteriorating factor of building materials
in the case of monuments and historic buildings [2]. The
saline deposits on the stone surface may be efflorescences,
subefflorescences or crusts [3].
Consolidation and protection conservation treatments
lead to the desired modification of stone characteristics
(microstructural and mechanical), preventing decay of stone
due to grains de-cohesion. Proper use of conservation materials leads to reduction of stone susceptibility to decay
forms as soluble salt crystallization and frost phenomena.
A lot of conservation materials used in the past for specific interventions proved to be improper, lacking chemical
relevance to the substrate and intensifying, this way, deterioration phenomena of the stone after their application [4].
Under supersaturated conditions, the growing crystals of
a salt solution exert pressure against the porewalls, known
Corresponding author.
E-mail address: amoropul@central.ntua.gr (A. Moropoulou).

0300-9440/$ see front matter 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S0300-9440(03)00110-3

as crystallization pressure [5,6]. According to Fitzner and

Snethlage idealized model [7], based on the thermodynamics of frost action in porous materials developed by Everett
[8], crystallization pressure is a function of the porous
structure of the stone. According to Gauri et al. [9] and
Rossi-Manaresi and Tucci [10], salts can become critical
agents of decay only when certain porosity characteristics
exist in the interior of the stone. Lewin [11] introduced
the mechanical characteristics of the stone in the investigation of crystallization pressure. The released free energy
from the spontaneous crystallization of a supersaturated
salt solution is compared to the mechanical strength of
the stone. However, particularities of stone and porous
size distributions were not taken into consideration in this
approach.
In the present work, the evaluation of consolidated and
protected building stones susceptibility to salt decay was
attempted by thermodynamics of crystallization in porous
materials. Calculations were realized taking into consideration both mechanical and microstructural characteristics of
the stone. NaCl was selected as a representative salt for the
calculation of crystallization pressure, due to its dangerous
action for the stone and its frequent appearance, especially
in marine environment.

260

A. Moropoulou et al. / Progress in Organic Coatings 48 (2003) 259270

During initial crystallization, salt crystals grow selectively

in the coarse pores of the stone [8]. When coarse pores
are filled with crystals, crystallization may advance either
disrupting the stone pore walls or filling successively the
fine pores [12].
The evaluation of the above thermodynamic alternatives
demonstrates the possible scenario for the development of
crystallization.
Artificial weatheringaging tests were also performed in
order to evaluate the effectiveness of the applied conservation materials.

2. Experimental procedure
Various conservation materials have been used in the
past, according to the demands of the case under examination. Inorganic materials have been widely used, due to
their chemical relevance to the authentic building material. Their use is confined nowadays, because they usually
form undesirable crusts and by-products after their application to the stone. Organic materials lead to the increase
of stones mechanical strength values. However, they tend
to form crusts and degrade during their exposure to oxygen
and ultraviolet radiation. Alkoxysilanes are the preferred
materials for the treatment of siliceous sandstones, due to
the formation of chemically relevant materials to the binder
during polymerization [4].
The examined Cyprus porous stone from Limassol quarry
was a calcareous sandstone, which consisted mostly of calcium carbonate (approximately 94%), with extremely small
amounts of quartz and plagioclases [4].
The consolidation material used was ethyl silicate dissolved in ethyl alcohol 20% (w/w) to SiO2 70% (w/w) in
the form of transparent liquid (codename RP) (Table 1).
It is a non-filmogenic material, penetrating, with chemical
relevance to siliceous sandstones.
The protective material consisted of a polysiloxane binder
dissolved in white spirit (codename GP). It is a colorless,
penetrating waterproofing agent for mineral or exterior surfaces, which adheres to the surface without forming a film
and affecting the vapor transmission properties of the stone.
A combination of the above consolidation and protective
materials was also applied (codename RG).
The porous stone samples were treated by the above
mentioned consolidation and/or protection materials by
immersion. The materials were applied until the weight of
Table 1
Ethyl silicate norm standards
Appearance

Transparent liquid

SiO2 (%, w/w)

Viscosity (25 C, cp)
Specific gravity (15 C, g/cm3 )
Flash point ( C)

70
3.3
0.970
17

the treated stone remained practically invariable, i.e. the

variation between the two last weights of the stone did not
exceed 5%.
Three kinds of treated samples were used, consolidated
(codename RPL), protected (codename GPL), consolidated
and then protected after 15 days (codename RGL). Regarding protection material, its drying time was 2 h and the whole
effect appeared after 24 h.
All types of samples along with the untreated ones (codename NTL) were submitted to the following experimental
tests:
Mercury intrusion porosimetry (performed by a Fisons
Porosimeter 2000) for the evaluation of microstructural
characteristics.
Compression and bending tests (performed by a Wykeham
Farrance Model 55662 Ho3 machine based on prEN 772-1
standard test method) for the evaluation of mechanical
characteristics.
Based on the obtained results, NaCl crystallization
pressure was calculated and the most probable development of the deterioration phenomenon, that is disruption
of the stone pore walls or filling of the smaller pores was
estimated.
Artificial weatheringaging tests of marine salt spray for
the evaluation of durability of the treated samples. The
tests included daily cycles of marine salt spray, pause and
drying. The cycle length of seawater spray on the stone
was 1 min, the flow rate 1020 cm3 /min and the received
salt quantity 5 mg seawater/cm2 of the stone surface. The
pause cycle was 1 min long, in order to allow the attachment or penetration of the salts on the stone surface. The
drying cycle was 28 min and air 40 C was sprayed on the
stone surface. The specimens remained in the simulation
chamber for 2 months and were subjected to weight and
ultrasonic velocity measurements during the tests.

3. Results and discussion

3.1. Mercury intrusion porosimetry
Mercury intrusion porosimetry measurements revealed
that the examined untreated stone presented values of total
porosity approximately 30%, active porosity 79% and average pore radius 34 m (Table 2). A great percentage of
pores presented radius size between 1 and 5 m (Table 3).
Results indicated that after the treatment there was an
alteration of the stones microstructural characteristics for
all types of specimens (Table 2). The ratio of active to total
porosity, which represents the percentage of pores communicating with the environment, was decreased compared
to the untreated specimens (Fig. 1). Average pore radius
was also decreased, indicating deposition of the conservation materials into the pores of the stone (Fig. 2). As
far as specific surface area is concerned, the values were

A. Moropoulou et al. / Progress in Organic Coatings 48 (2003) 259270

261

Table 2
Microstructural characteristics data for untreated and treated porous stone by mercury intrusion porosimetry
Sample

Total cumulative
volume (mm3 /g)

Specific surface
area (m2 /g)

Average pore
radius (m)

Bulk density
(g/cm3 )

Total
porosity (%)

Active
porosity (%)

NTL1
NTL2
NTL3
NTL4
RPL1
RPL2
GPL1
GPL2
RGL1
RGL2

142.97
156.79
158.76
158.24
114.14
108.54
119.00
103.65
92.29
117.85

1.06
1.30
1.16
1.15
0.76
0.73
0.99
1.11
0.86
0.73

3.25
3.00
4.06
3.39
2.17
1.99
3.56
2.05
1.22
2.72

2.08
1.96
1.94
1.95
2.06
2.13
2.10
2.14
2.21
2.10

29.73
30.73
30.80
30.85
23.51
23.11
24.99
22.18
20.39
24.74

7.11
8.50
9.37
9.78
4.47
3.97
5.89
4.25
3.23
4.21

Table 3
Pore percentage volume data for untreated and treated porous stone by mercury intrusion porosimetry
Sample

NTL1
NTL2
NTL3
NTL4
RPL1
RPL2
GPL1
GPL2
RGL1
RGL2

Pore percentage volume, V

r1 = 00.01 m

r2 = 0.010.1 m

r3 = 0.10.5 m

r4 = 0.51 m

r5 = 15 m

R1 = 510 m

R2 = 10100 m

1.38
1.95
1.51
1.62
0.90
0.71
1.38
1.79
1.14
1.03

7.10
7.82
7.76
7.29
4.88
5.35
6.08
5.94
5.11
4.37

5.54
5.65
5.75
5.88
6.67
6.54
4.88
6.66
7.74
5.14

5.13
5.31
5.33
6.23
8.23
8.74
5.75
8.78
11.26
5.94

112.23
122.02
122.74
123.94
86.89
81.33
90.03
73.64
60.44
94.33

6.10
8.93
9.38
6.86
2.90
2.98
4.95
3.39
2.70
3.27

5.50
5.11
6.29
6.43
3.67
2.88
5.93
3.46
3.90
3.77

reduced after the treatment, indicating a decrease in the

percentage of fine pores (Fig. 3). Reduction of fine pores
is desirable. Water in fine pores tends to be absorbed and
retained for long periods. Therefore, condensation is more
likely to occur in them [13]. In the same way, bulk density

35

values presented an increase after the treatment, indicating

amelioration of stones cohesion (Fig. 4).
The phenomena were more intense for the consolidated
samples compared to the protected ones, due to the different activity between conservation materials. The deposition

31.7
30.4
27.7

30
23.9

23.6

AP/TSP [%]

25
19.2

19.0

20

17.2
15.8

17.0

15

10

0
NTL1

NTL2

NTL3

NTL4

RPL1

RPL2

GPL1

GPL2

RGL1

RGL2

Sample
Fig. 1. Active to total porosity ratio values, AP/TSP, for untreated and treated porous stone.

262

A. Moropoulou et al. / Progress in Organic Coatings 48 (2003) 259270

4.5

4.06

4
3.5

3.56
3.39

3.25
3.00

2.72

PRA [m]

3
2.17

2.5

2.05

1.99

2
1.22
1.5
1
0.5
0
NTL1

NTL2

NTL3

NTL4

RPL1

RPL2

GPL1

GPL2

RGL1

RGL2

Sample
Fig. 2. Pore radius average values, PRA, for untreated and treated porous stone.

1.4

1.3
1.16

1.2

1.15

1.11

1.06
0.99

0.86

SSA [m 2 /g]

0.76

0.73

0.73

0.8
0.6
0.4
0.2
0

NTL1

NTL2

NTL3

NTL4

RPL1

RPL2

GPL1

GPL2

RGL1

RGL2

Sample
Fig. 3. Specific surface area values, SSA, for untreated and treated porous stone.

mechanism of a consolidant leads to more pronounced

changes of microstructural characteristics of the stone,
whereas a protection material leads to hydrophobic behavior of the stone. In general, results could indicate a probable
trend for lower susceptibility of stone to salt decay.

of the stone after the treatment. The examined stone was

quarried and therefore did not present serious signs of decay
due to exposure to the environmental conditions.
Moreover, bending strength measurements were performed. The corresponding bending strength values were used

3.2. Mechanical tests

Table 4
Compression strength measurements data for untreated and treated porous
stone

Compression strength measurements revealed that the

stones strength was increased after the treatments and
especially after protection (Table 4). According to previous
works [14,15], such an increase could be considered insignificant, avoiding any serious problems caused by great
and incompatible modifications of the mechanical properties

Sample

c (N/cm2 )

NTL
RPL
GPL
RGL

4040
4598
5066
4302

A. Moropoulou et al. / Progress in Organic Coatings 48 (2003) 259270

2.25

263

2.21

2.2
2.14

2.13

2.15

2.10

2.10
2.08

BD [g/cm 3 ]

2.1

2.06

2.05
2

1.96
1.94

1.95

1.95
1.9
1.85
1.8
NTL1

NTL2

NTL3

NTL4

RPL1

RPL2

GPL1

GPL2

RGL1

RGL2

Sample
Fig. 4. Bulk density values, BD, for untreated and treated porous stone.

for the calculation of the tensile strength of the specimens

(Table 5). This is due to the fact that tensile strength measurements are difficult to materialize, because samples of
great length are required. In the case under examination, this
could not be possible. The results are in accordance with
those of the compression strength measurements, i.e. protection treatment leads to the greatest increase of the mechanical strength of the porous stone.
3.3. Crystallization pressure
Crystallization of soluble salts is the main reason of
weathering of building stones. Salts migrate in the form of
solution through the complex capillary system of the stone
towards the surface and may crystallize as efflorescences
on the surface of the stone under favorable conditions or
subefflorescences below it. This latter phenomenon may
cause stones decay [3].
When salt solution migration to the surface of the stone
is faster than the rate of the drying, the crystals of solute
deposit on the external surface and efflorescences appear.
On the contrary, when salt solution migration to the surface
of the stone is slower than the rate of the drying, the water
evaporates, the solute crystallizes into the pores and subefflorescences are formed, without any visible modification of
the stone [10].
Table 5
Tensile strength measurements data for untreated and treated porous stone
Sample

t (N/cm2 )

NTL
RPL
GPL
RGL

958
1090
1201
1020

In the last case, during NaCl crystal growth in the coarse

pores of the stone, an excess pressure, P, is developed. If
this pressure exceeds the strength of the porous material,
mechanical failure occurs. Crystallization pressure, P, as a
function of the pore dimensions, is given by Eq. (1) [8]:



1
1
P = 2

(1)
ri
R
where P is the crystallization pressure, the interfacial crystal saltsolution tension, ri the sum of small pores radii of
the porous stone, and R the radius of coarse pore of the
porous stone.
In the case of the examined stone, pores with radii greater
than 5 m represent a significant percentage of the pore volume, as indicated by mercury intrusion porosimetry results
(Table 3). In such a case, preferential crystal growth takes
place in them [10].
As a consequence, in Eq. (1) the factor 1/R can be eliminated and Eq. (2) can be obtained:


1
P = 2
(2)
ri
The effective crystallization pressure for each class of pore
size is determined multiplying the theoretical value with
the factor Vri /VR determined by the porosimetric measurements, where Vri is the total percentage of the volume of
pore size ri and VR the total percentage of the volume of
large pores (R > 5 m) [16].
Crystallization pressure values were increased after the
treatment (Table 6, Fig. 5). These values cannot be considered valid and in accordance with the anticipated results
after treatments application. This is probably due to the fact
that the examined porous stone is inhomogeneous and consists of pores, which cannot be simulated by a cylindrical
model in a realistic way.

264

A. Moropoulou et al. / Progress in Organic Coatings 48 (2003) 259270

Crystallization pressure values are correlated with the

corresponding compressive strength values, c , of the untreated and treated stone (Fig. 6). Crystallization pressure
values are significantly lower than those of the compressive strength in each case, showing that treatments did
not cause significant changes on the stone characteristics. In the same way, crystallization pressure values are
compared to the experimentally measured tensile strength
values.
In engineering materials the tensile strength, t , is the
quantity, which on a macroscopic level reflects the cohesive
strength of the solid, presenting the maximum load per surface area, which can be supported without fracture. When

Table 6
Crystallization pressure values for untreated and treated porous stone
Sample

P (N/cm2 )

NTL1
NTL2
NTL3
NTL4
RPL1
RPL2
GPL1
GPL2
RGL1
RGL2

708.15
743.28
571.66
692.63
889.67
910.08
1017.51
1327.72
1013.01
853.06

1327.7

1400
1200

1013.0

1017.5

1000

P [N/cm ]

743.3

800

889.7

910.1

RPL1

RPL2

853.1

692.6

708.2
571.7

600
400
200
0

NTL1

NTL2

NTL3

NTL4

GPL1

GPL2

RGL1

RGL2

Sample
Fig. 5. Crystallization pressure values, P, for untreated and treated porous stone.

6000

P- c [N/cm 2 ]

5000

4000

3000

2000

1000

NTL1

NTL2

NTL3

NTL4

RPL1

RPL2

GPL1

GPL2

RGL1

RGL2

Sample
P

Fig. 6. Comparison between crystallization pressure values, P, and compressive strength values, c , for untreated and treated porous stone.

A. Moropoulou et al. / Progress in Organic Coatings 48 (2003) 259270

265

1400
1200

P- t [N/cm 2 ]

1000
800
600
400
200
0

NTL1

NTL2

NTL3

NTL4

RPL1

RPL2

GPL1

GPL2

RGL1

RGL2

Sample
P

Fig. 7. Comparison between crystallization pressure values, P, and tensile strength values, t , for untreated and treated porous stone.

the crystallization pressure surpasses the tensile strength, the

material is damaged (Fig. 7) [17].
Crystallization pressure of Limassol stone generally presented lower values than those of tensile strength. It is
remarkable, though, that the tensile strength values were of
about the same magnitude of the calculated crystallization
pressure values, especially in the case of treated specimens.
This fact intensifies the estimation that the tensile strength
is more representative for the description and simulation of
the real phenomenon than the compressive strength.
3.4. Thermodynamic approach of crystallization
According to the first law of thermodynamics, it may be
stated that internal expansion due to crystallization pressure
can occur provided there is a process offering the consumed energy. This process, which balances the consumed
energy by the release of free energy, could be the spontaneous deposition of crystals from a supersaturated solution
[11].
Crystal growth comprises ion transfer from the salt solution to the surface of the crystal in combination with
reactions on the surface, which lead to the incorporation of
the ion into the crystal lattice. The slowest phenomenon is
the rate-determining step of crystal growth.
For the precipitation of a single crystal from the solution
the Eq. (3) is valid [18]:

Gn =
Gbulk +
Gsurf

(3)

where
Gn is the free energy of formation of the crystal
(subscripts bulk and surf refer to the bulk and interfacial
contributions to the free energy, respectively).
The free energy that corresponds to progressive crystal
growth, pressure increase and mechanical failure,
G1 , is

estimated by expressions (4) and (5) [7,8]:

G1 =
Gbulk +
W



IAP

Gbulk = nRT ln
K

(4)
(5)

where [16] IAP is the ion activity product in the original

supersaturated solution, which determines whether a salt can
precipitate or not, K the equilibrium ion activity product,
and n the number of moles precipitated, and
IAP = [Na+ ][Cl ] K

(condition for supersaturation)(6)

Considering that the developed stress in the interior of the

stone due to crystallization process is compressive, the work
done is approximated by the average pressure during crystallization process, which is equal to 1/2 of the final pressure
[11]:

W = 21 c
V

(7)

where
W is the work to be done for the disruption of
the pore walls,
V the volume change when the stone is
submitted to a stress equal to the compressive strength, and
c the compressive strength of the stone.
When the stone is squeezed by a stress c , it is compressed
to the extent [11]:
k=

1
V
c V

(8)

where k is the bulk modulus of compressibility, V the initial

of the stone specimen [19].
The free energy, which corresponds to propagation of the
salt crystals along the capillaries,
G2 , after filling of coarse
pores, is [7,8]

G2 =
Gbulk +
Gsurf

(9)

266

A. Moropoulou et al. / Progress in Organic Coatings 48 (2003) 259270

where
Gsurf is the surface change contribution to free
energy of NaCl crystal formation (=
A), the specific
interfacial energy between the crystal and aqueous solution,
and
A the change of the surface area of the crystal.
The internal surface area, which corresponds to a volume

V, is calculated using the spherical model for the pores

shape:

i 

A = 3
V
(10)
ri

Table 7
Free energy difference for untreated and treated porous stone as a function
of compressive strength

where i is the percentage of the volume of pore size ri .

Considering that the tensile strength is representing the
developed stress in the pores of the stone, mechanical failure
takes place when the elastic limit of the stone is surpassed.
In this case, the work done is approximated assuming the
average pressure equal to 1/2 of the final pressure, which
is the tensile strength of the stone. In this case, the representative factor is the modulus of elasticity instead of the
modulus of compressibility:
t
l
V
E=
= t
= t

Sample

G1
G2 (N/cm2 )

NTL1
NTL2
NTL3
NTL4
RPL1
RPL2
GPL1
GPL2
RGL1
RGL2

280.93
278.61
280.91
280.71
368.08
368.67
443.72
437.47
315.88
321.43

Table 8
Free energy difference for untreated and treated porous stone as a function
of tensile strength

(11)

where E is the modulus of elasticity, t the tensile strength

of the stone, the strain,
l the length change when the
stone is submitted to a stress equal to tensile strength, and

V the volume change when the stone is submitted to a

stress equal to tensile strength.
For crystals larger than 5 m, the bulk free energy Gbulk
is high enough so that
Gsurf can be neglected in thermodynamic calculations.
However, during crystallization into pores smaller than
5 m, surface free energy is an important factor of crystal
growth.
In order to evaluate the most probable thermodynamic
scenario, calculations and comparisons between
W and

Sample

G1
G2 (N/cm2 )

NTL1
NTL2
NTL3
NTL4
RPL1
RPL2
GPL1
GPL2
RGL1
RGL2

1.03E01
9.83E02
1.02E01
1.02E02
1.40E01
1.41E01
1.65E01
1.53E01
1.11E01
1.21E01

Gsurf have been performed, considering compressive and

tensile strength, respectively, as representative parameters of
the stone cohesion. From Eqs. (4) and (9):

G1
G2 = (
Gbulk +
W) (
Gbulk +
Gsurf )
=
W
Gsurf
443.7

450
368.1

400

437.5

368.7
315.9

350

G 1 -G 2 [N/cm 2 ]

280.9

300

278.6

280.9

(12)

321.4

280.7

250
200
150
100
50
0

NTL1

NTL2

NTL3

NTL4

RPL1

RPL2

GPL1

GPL2

RGL1

RGL2

Sample
Fig. 8. Free energy difference values,
G1
G2 , of untreated and treated porous stone as a function of compressive strength.

A. Moropoulou et al. / Progress in Organic Coatings 48 (2003) 259270

1.80E -01

267

1.7E -0 1
1.5E -0 1

1.60E -01

1.4E -0 1

1.4E -0 1

1.40E -01
2

1.2E -0 1
1.0E -0 1

1.20E -01

G 1 -G 2 [N/cm

1.0E -0 1

9.8E -0 2

1.1E -0 1

1.0E -0 1

1.00E -01
8.00E -02
6.00E -02
4.00E -02
2.00E -02
0.00E +00

NTL1

NTL2

NTL3

NTL4

RPL1

RPL2

GPL1

GPL2

RGL1

RGL2

Sample
Fig. 9. Free energy difference values,
G1
G2 , of untreated and treated porous stone as a function of tensile strength.

Thus, when
W >
Gsurf , propagation of the salt crystals along the capillaries is the thermodynamically preferred
process, whereas when
W <
Gsurf , it is the growth of
bulk NaCl crystals and mechanical failure.
In all types of specimens under examination the released
free energy during the spontaneous crystallization from a
supersaturated solution was compared to the mechanical energy needed for the materials disruption in order to evaluate
the most probable thermodynamic scenario (salt crystallization or rupture).
The examined samples presented higher
W values than
the
Gsurf ones (
G1 >
G2 ) (Tables 7 and 8, Figs. 8
and 9) indicating thus that the most probable scenario was

the propagation of salt crystals along the capillaries and the

filling of the stone smaller pores. This suggestion stands
for both treated and untreated samples, showing that the
Limassol stone is highly resistant to decay forms caused by
soluble salts. The comparison of the relative values reveals
that this scenario is more possible for treated specimens.
Using tensile strength values as a representative factor for
stone cohesion, the obtained free energy results correspond
to progressive crystal growth and filling of smaller pores.
The free energy values were positive (>0) but close to zero
(Table 8, Fig. 9). This fact could be accepted as an indication that the probable thermodynamic scenario was not
clear.

5
4.5
4

NTL
Final Weight Gain:4.41%

Weight Gain [%]

3.5
3
2.5
Final Weight Gain: 0.4-0.5%

2
Final Weight Gain:0.25-0.29%

1.5

RGL2

Final Weight Gain:0.28%

RG L1

RPL2

RP L1

GPL

1
0.5
0
0

9 11 14 16 18 21 23 25 28 30 32 35 37 39 42 44 50 56 60

Days of exposure
Fig. 10. Weight variation values of untreated and treated porous stone as a function of days of exposure.

268

A. Moropoulou et al. / Progress in Organic Coatings 48 (2003) 259270

3.5. Artificial weathering tests of marine salt spray

Marine spray simulation tests revealed the durability under aggressive conditions of the porous stone after the treatment. Studying the time dependence of the weight variation
of stone exposed to these conditions, significant reduction of
the weight variation percentage was observed for the treated
samples (Table 9, Fig. 10). During the 60-day salt spray simulation, the weight gain was minimum (0.250.5%), indicating reduction of salt accumulation phenomena on the stone
surface. Moreover, ultrasonic velocity was measured in three
dimensions of the prismatic samples because of the stone

Table 9
Weight variation of treated and untreated samples during artificial weathering tests
Sample

Final weight gain (%)

NTL
RPL
GPL
RGL

4.41
0.40.5
0.28
0.250.29

anisotropy (Figs. 1113). The resulted values did not reveal

significant fluctuations, showing maintenance of the mechanical strength of the stone and physico-mechanical compatibility between the stone and the conservation materials.

4600
NTL

Ultrasonic velocity (m/sec)

4500

4400
RPL2

GPL

4300

RGL2

4200

RPL1

4100
RGL1
4000
0

9 11 14 16 18 21 23 25 28 30 32 35 37 39 42 44 50 56 60

Days of exposure
Fig. 11. Ultrasonic velocity values (v variation) of untreated and treated porous stone as a function of days of exposure.

4700
NTL

Ultrasonic velocity (m/sec)

4500
RGL2

4300
RPL2
GPL

4100

3900

3700

RPL1

RGL1

3500
0

9 11 14 16 18 21 23 25 28 30 32 35 37 39 42 44 50 56 60

Days of exposure
Fig. 12. Ultrasonic velocity values (v1 variation) of untreated and treated porous stone as a function of days of exposure.

A. Moropoulou et al. / Progress in Organic Coatings 48 (2003) 259270

269

4600
GPL

4400
RGL1

Ultrasonic velocity (m/sec)

4200
4000
RPL1

3800
RGL2
NTL

RPL2

3600
3400
3200
3000
0

9 11 1 4 16 18 2 1 23 2 5 28 3 0 3 2 35 3 7 39 42 4 4 5 0 56 6 0

Days of exposure
Fig. 13. Ultrasonic velocity values (v2 variation) of untreated and treated porous stone as a function of days of exposure.

4. Conclusions
Susceptibility of stone to salt decay is a function of mechanical and structural parameters, i.e. compressive strength,
modulus of compressibility, tensile strength, modulus of
elasticity, pore size distribution. Consolidation treatments
aim to the modification of the above characteristics of the
stone, in order to achieve reduction of the decay phenomena
caused by soluble salts.
In the present work, porous stone from Cyprus was
consolidated and protected with certain conservation materials. From the experimental procedure the following is
obtained:
Mercury intrusion porosimetry measurements revealed
microstructural modifications and a porous shifting towards smaller pore radius, due to the treatments. These
phenomena were more intense for the consolidated than
the protected samples. Results indicated lower susceptibility of the examined stone to salt decay after the
treatment.
Compression and tensile strength values both of treated
and especially of protected samples were increased
compared to that of the untreated ones. The combined
treatment led to the slightest increase. All changes were
considered to be within compatible limits.
Crystallization pressure values, which were increased
after the treatments and especially after protection, could
not be considered valid in the case of the specific stone,
due to its inhomogeneity. Comparisons between the crystallization pressure and either, the compressive strength
or the tensile strength showed that mechanical failure of
the treated stone did not seem to be probable.

Thermodynamic analysis showed that the most probable

scenario after the salt precipitation in the stone was filling of the smaller pores and not disruption of the stone
pore walls. It should be noted, though, that calculations
based on the tensile strength of the stone resulted in extremely small values of free energy difference. That is
why, the spontaneous phenomenon of crystallization was
not clearly determined in this case. Protected samples
presented the highest free energy difference values, indicating the strongest probability of the above mentioned
thermodynamic scenario validity.
Conservation materials succeeded to reduce salt accumulation phenomena, which occur during the marine spray
simulation tests. Weight variations of treated samples
presented the lowest values.
The Limassol porous stone was found to present high
mechanical strength values, which were intensified by the
conservation treatments, indicating physico-mechanical
compatibility between the stone and the conservation
materials. Total evaluation of results indicated that the
combined conservation treatment gave the best results.
Acknowledgements
Thanks should be attributed to architect Mr. Tzanetakis,
Technical Director of the National Bank of Greece, for
the support and permission to perform the above investigation with the purpose to advise technically the construction
of the new building in Aiolou and Sophokleous street in
Athens.
Furthermore, acknowledgements are also attributed to
Elli Vintzileou, NTUA Professor, for the performance of

270

A. Moropoulou et al. / Progress in Organic Coatings 48 (2003) 259270

the mechanical tests in the Reinforced Concrete Laboratory

in the National Technical University of Athens, Greece.
[10]

References
[1] S.D. Leith, M.M. Reddy, W.F. Ramirez, M.J. Heymans, Environ. Sci.
Technol. 30 (1996) 2202.
[2] A. Arnold, in: F. Zezza (Ed.), Proceedings of the First International
Symposium for the Conservation of Monuments in the Mediterranean
Basin, Bari, 1989, p. 59.
[3] V. Lopez-Acevedo, C. Viedma, V. Gonzalez, A. La Iglesia, J. Cryst.
Growth 182 (1997) 103.
[4] A. Moropoulou, N. Kouloumbi, A. Bakolas, G. Haralampopoulos,
Pitture e Vernici Eur. Coat. 1213 (2001) 19.
[5] P. Theoulakis, A. Moropoulou, in: J. Ciabach (Ed.), Proceedings of
the Sixth International Congress on Deterioration and Conservation
of Stone, Torun, 1988, p. 86.
[6] A. Moropoulou, P. Theoulakis, in: D. Decrouez, J. Chamay, F. Zezza
(Eds.), Proceedings of the Second International Symposium for the
Conservation of Monuments in the Mediterranean Basin, Geneva,
1991, p. 493.
[7] B. Fitzner, R. Snethlage, GP News Lett. 3 (1982) 13.
[8] D.H. Everett, Trans. Faraday Soc. 57 (1961) 1541.
[9] K.L. Gauri, A.N. Chowdhury, N.P. Kulshreshtha, A.R. Punuru,
in: Marinos, Koukis (Eds.), Engineering Geology of Ancient

[11]

[12]
[13]
[14]

[15]

[16]

[17]
[18]
[19]

Works, Monuments and Historical Sites, Balkema, Rotterdam, 1988,

p. 723.
R. Rossi-Manaresi, T. Tucci, in: F. Zezza (Ed.), Proceedings of the
First International Symposium for the Conservation of Monuments
in the Mediterranean Basin, Bari, 1989, p. 97.
S.Z. Lewin, in: F. Zezza (Ed.), Proceedings of the First International
Symposium for the Conservation of Monuments in the Mediterranean
Basin, Bari, 1989, p. 59.
P. Theoulakis, A. Moropoulou, Construct. Building Mater. 11 (1997)
65.
D. Camuffo, Water Air Soil Pollut. 21 (1983) 151159.
R.L. Gauri, J.A. Gwinn, R.K. Popli, Proceedings of the International
Symposium on the Deterioration of Building Materials, Athens,
Greece, 1976, p. 143.
J.R. Clifton, Preservation of Historic Adobe StructuresA Status
Report, NBS Technical Note 934, National Bureau of Standards,
Washington, DC, 1977.
K.L. Gauri, R. Adinarayana, R. Punuru, in: F. Zezza (Ed.), Proceedings of the First International Symposium for the Conservation of
Monuments in the Mediterranean Basin, Bari, 1989, p. 255.
A. La Iglesia, V. Gonzalez, V. Lopez-Acevedo, C. Viedma, J. Cryst.
Growth 177 (1997) 111.
R.A. Berner, Early Diagenesis: A Theoretical Approach, Princeton
University Press, New York, 1980, p. 91.
P. Theoulakis, A. Moropoulou, in: Proceedings of the Fourth
International Symposium on the Conservation of Monuments in the
Mediterranean Basin, vol. 4, Rhodes, 1997, p. 647.