Construction and Building Materials 23 (2009) 3321–3327

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

The role of salt fog on alteration of dimension stone

Z.S.G. Silva *, J.A.R. Simão
Centro de Investigação em Ciência e Engenharia Geológica, Departamento de Ciências da Terra, FCT, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal

a r t i c l e i n f o a b s t r a c t

Article history: Limestones and silicate rocks used as dimension stone were submitted to cycles of salt fog, totalizing
Received 24 July 2008 140 days in a salt spray chamber. Rock samples comprised three types of limestone, three different mar-
Received in revised form 16 June 2009 bles, six types of granites, two syenites, one gabbro and one basalt. Mineral changes and variation on sur-
Accepted 18 June 2009
face luster, weight loss and salt crystallization on rock surfaces were some of the features observed along
Available online 17 July 2009
the exposure. While losses of weight and surface luster were the most striking features observed on car-
bonate rocks, mineral degradation was the main modification on silicate rocks, in which biotite showed
Keywords:
to represent a key mineral in the alteration process due to its cleavage and possibly to its composition.
Salt fog
Rock alteration
Ó 2009 Elsevier Ltd. All rights reserved.
Degradation
Dimension stone

1. Introduction Salt fog activity comprises mechanisms such as salt crystalliza-

tion and deposition of salt on the rock surface, salt solution pene-
Salt fog is a powerful rock weathering agent which has been tration in rock pores and further salt expansion causing the
shown to effectively act on rocks of any composition. The first signs increase of voids in the rock framework, followed by removal of
of alteration appear on the rock surfaces, even on silicate rocks soluble salts and secondary minerals. Several authors have pointed
thought in the past to be a very resistant, long lasting material. out the various ways in which salt weathering affects different
Changes on the rock structure follow as the salt action goes on. types of rocks [1,2] and mechanisms acting on rocks [3].
Historic buildings and monuments showing signs of degrada- The wide variety of silicate rocks used as dimension stone con-
tion have called the attention of researchers and technicians in- stitutes an important field of research. It opens the opportunity of
volved in cleaning, conservation and restoration of such heritage. observing rocks of different origins and mineral compositions
In this sense, due to their significance to society, they should not answering in different ways to the action of salt fog. Following this
be lost. The methodology in which any relevant work can be based idea, different rock types used in buildings, either having structural
on for recovery of those buildings is a complete knowledge of the function or having finishing purposes on façades, were chosen and
material (rock), its chemistry and mineral composition and there- submitted to salt spray cycles. These rocks were observed during
after, the knowledge of the causes of alteration, their extent and several cycles and a comparison between the material just prior
how to remediate the problem. This implies, in many instances, to exposition and after each salt fog cycle could be made recording
an operation that starts with the removal of the cause, if possible, changes on the rock surfaces and on the rock as a whole. The find-
followed by the repair of the damaged object (building). Among the ings do not represent the only alteration of each material, but they
several types of rock alteration, salt activity inside the rock is one certainly show evidence of possible ‘‘starters” or ‘‘promoters” of a
common factor that plays an important role on rock decay. It is certain type of alteration for those rocks.
necessary then, to get to know the relationship between the origi- Rocks chosen for the experiments are commonly used as
nal rock and new materials added to it, a process developed in dimension stone in Portugal. They come from quarries having
many circumstances. The fact that salt activity is not an isolated different geological origins, either in Portugal or in Brazil, being
feature restricted to the surface leads to the need of understanding all of them of common use as building materials and dimension
the way it reaches the rock, how it acts on the rock surface and the stone.
type of exchange which takes place within the rock, between
components. Buildings exposed to salt atmosphere near the ocean 2. Methodology
usually give some hints to researchers what lead to diagnose of the
occurrence and the proposal of solutions for remediation. 2.1. Samples selection

Limestones and marbles, granites, syenites, gabbros and basalt were tested in
* Corresponding author. Tel.: +351 212948573; fax: +351 212948556. the laboratory through the exposition to salt fog during 13 cycles of 10 days each.
E-mail address: zcs@fct.unl.pt (Z.S.G. Silva). Intentionally: (1) limestones of different porosities and (2) silicate rocks containing

0950-0618/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2009.06.044
3322 Z.S.G. Silva, J.A.R. Simão / Construction and Building Materials 23 (2009) 3321–3327

stable minerals like quartz, in granites, unstable minerals such as nepheline, soda- 2.5. Observations and changes on rock surfaces
lite and cancrinite in syenites and olivine, in gabbro were selected. Most silicate
rocks contain biotite as a common accessory mineral, which was identified as a Sample surfaces were observed after each cycle and most of them were docu-
key mineral in the process of rock disaggregation and alteration [4]. mented with pictures. In general, siliceous rocks have low porosity, although differ-
ent framework and mineral heterogeneity cause differences in behaviour regarding
2.2. Experiments with salt spray chamber internal liquid circulation, a fact which could not be interpreted taking into account
just one single cause. Common features like grain contour limits, cleavage traces
Samples weighed between 23.57 and 106.55 g, were 3 cm  9 cm  1 cm in size and fractures on the rock surface were noted to be privileged sites for salt deposi-
and all had one polished surface which was set upwards in the camera. All samples tion and later expansion.
were dried and weighed before being submitted to the salt spray. Data will be dis- Carbonate rocks have a wider diversity in rock behaviour; marbles usually have
played on the appropriate section. low porosity due to closing pores during metamorphism, but on the contrary, lime-
The salt spray chamber (Ascott S120T) was fed with a salt solution (180 g of stones can show a wide range of porosity values as consequence of their varied ori-
NaCl per liter of water). Each spray cycle comprised five turns of 12 h of active spray gin and pore size distribution. As a result, alteration patterns are diverse, such as
and 36 h rest, totalizing 10 days of exposure. After each 10 days cycle the samples sanding, fractures and alveolar weathering. Rocks discussed in this paper have dif-
were taken to a stove at 60 °C, and after 24 h they could have their surfaces ob- ferent framework, varying from microcrystalline to oolitic to mixed structure, all
served and documented with photographs. Samples were washed with running features having an important role in the fluid infiltration and on the formation of
and de-ionized water during 5 days after which they were dried and then weighed. soluble salts in the rock interior.
During the first salt fog cycles salt deposition occurred in most surfaces, but
after the 4th cycle, specific features became very distinct. In silicate rocks fracture
2.3. Petrography of siliceous rocks
and cleavage traces in feldspars and micas and grain contours were filled with salt
fine crystals. As the experiment proceeded, mica flakes started coming out of the
Granites: Monforte (MonfGr), Monção (MonçGr), Castelo de Vide (CVGr), Blue
rock, a feature to be intensified during the 9th and 11th cycles in which the flakes
Socorro (BlGr), Sound Vila Real (SVRGr) and Altered Vila Real (AVRGr). All granites
opened along the cleavage planes above the rock surface. Monção, Monforte, Cast-
have a medium to coarse granular texture. Blue Socorro is a slightly deformed gran-
elo de Vide and Blue granite show the best examples of such behaviour. At the same
ite with a blue quartz as distinct feature. The two samples from Vila Real granite
time, plagioclase grains exhibited signs of corrosion and alteration while on the sur-
were tested in order to identify any striking behaviour as they both come from
face of Vila Real (sound) granite muscovite and biotite flakes started opening along
the same quarry but are used with different purposes: the altered granite (from
their cleavage planes above the rock surfaces. Blue sodalite and nepheline syenites
the upper portion of the quarry) has been used in buildings in which an old aspect
and also Odivelas gabbro on the contrary, had their surfaces sprinkled with some
is desired to simulate old construction; the sound rock, exploited beneath the al-
spectacular halite crystals, exposing unique bell shaped and hopper crystals on
tered upper section of the massif has been used as an aggregate for concrete prep-
their polished surfaces. Hopper crystals result from salt deposition more rapid at
aration and other purposes.
the edges than at the centre, on smooth surfaces, giving rise to a funnel shaped
Syenites: Nepheline syenite (NeSy) from Monchique, Portugal and sodalite sye-
form, as exposed in most of the newly formed halite crystals. On the surface of
nite (SoSy) from Bahia, Brazil. Both rock types contain cancrinite as a minor accessory
Odivelas gabbro, fine halite crystals are frequent around the main, larger hopper
mineral, which is believed to be one of the poor silica minerals to start the alteration
crystal.
process in these rocks [5]. Nepheline and sodalite then follow the alteration path.
The various types of crystal patterns on the rock surfaces just described are
Gabbro: Odivelas gabbro (OdivGb) has a special meaning in this research since
illustrated in Figs. 1 and 2. The different crystal patterns were compared to what
the rock was widely used as dimension stone up to the point the quarry reached a
has been recorded in the literature [12,13]. Shapes here are different, the mineral
fractured zone and this fact implied in quick alteration of the quarried rock. The
is just one, NaCl, from the aerosol. The fact that there appear smaller sized crystals
presence of olivine played an important role on that alteration process due to the
around well formed hopper crystals could be accounted for different growth rate
fluid circulation trough the rock body and to the high susceptibility to alteration
during the 36 h rest of the salt cycle, implying variable crystallizing conditions;
of that mineral. Similar rocks containing olivine come up with the same type of
nucleation at that time could have been different as well.
alteration pattern [6,7].
Crystal shapes which are observed in other samples may result from factors
Basalt: Tuffaceous basalt from Azores (AzB). The rock is highly porous and its
such as variation on the nature of the rocks substratum (different mineral compo-
use as dimension stone has grown very rapidly in the last decade. There are some
sition, cleavage traces, wet rock surface, etc.). Needle shaped crystals which are re-
different types of volcanic rock of basaltic composition in the island and this sample
ported as common in other situations are not observed here. It is understood,
is just representative of this specific type of lava. Similar rocks from Madeira Island
though, that the two main parameters involved in the crystal formation, followed
have been studied in detail due to their large local use [8].
by rock alteration are both chemical and physical related, as pointed out elsewhere
Data on petrography and porosity of siliceous rocks [9–11] are given in Table 1.
[14]. It is useful to remind that rocks shown here have mineral structures varying
from nesossilicates (olivine) to chain (pyroxenes and amphiboles), phyllossilicates
2.4. Petrography of carbonate rocks
(micas) and tektossilicates (quartz, feldspars, fedspathoids), exposing cleavage
traces or not on the rock surface and this is also a factor to take into account during
Limestones: Oolitic limestone (Ool ls) from Alcobaça, Encarnadão (Red ls) and
salt solution deposition and crystal growth.
Negrais (Negr ls) from Sintra, Portugal. The first rock is very popular and is widely
Observations on carbonate rocks surfaces are distinct from the silicate rocks:
used in the country as dimension stone due to its uniform, light colour and homo-
apparently none of the marble samples seem to be affected by the salt fog. Salt pre-
geneous texture. The occurrence is widespread and may present slight differences
cipitates and deposits very fine crystals on the polished surfaces which as the
locally giving rise to a variety of commercial designations by different industries.
experiment proceeds, exhibit slight loss of brightness. However, regarding the lime-
The last ones (Red ls and Negr ls) are facies variations of the Lioz limestone from
stone, many changes take place as penetration of the chloride solution is very effec-
Sintra County.
tive due to their high porosity. All samples loose the brightness of the polished
Marbles: White marble (Wm), from Vila Viçosa, Pink marble (Pkm) and Black
surface very early. Nearly at the 4th cycle, oolitic limestone starts showing individ-
marble, Ruivina (Blkm), from Borba, Évora District. Data on petrography and poros-
ualization of oolites, at the same time that both Negrais and Encarnadão expose
ity [10] are shown in Table 2.
their rock fractures and their fossil content as individualized components. From

Table 1
Petrography and porosity of siliceous rocks.

Sample Texture Mineralogy Open porosity (%)

MonfGr Medium grained, homogeneous Quartz (22%), microcline (40%), Na-plagioclase (20%), biotite (13%), others (5%) 0.46
MoncGr Porphyritic, medium to coarse grained Quartz (25%), microcline (35%), Na-plagioclase (27%), biotite (10%), others (3%) 0.77
CVGr Porphyritic, medium to coarse grained Quartz (32%), microcline (36%), Na-plagioclase (16%), biotite (12%), muscovite (2%), 0.63
others (2%)
BIGr Coarse grained, slightly oriented Quartz (30%), microcline (35%), plagioclase (15%), hornblende (10%), biotite (6%), others (3%) 0.46
SVRGr Medium to coarse grained, slightly porphyritic Quartz (28%), microcline (32%), plagioclase (25%), biotite (3%), muscovite (10%), others (2%) 0.90
AVRGr Medium to coarse grained, slightly porphyritic Quartz (28%), microcline (32%), plagioclase (25%), biotite (3%), muscovite (10%), others (2%) 1.94
NeSy Medium to coarse grained Orthoclase (45%), nepheline (22%) plagioclase (2%), aegirine–augite (10%), biotite (6%), 0.89
others (8%)
SoSy Medium grained Microcline (50%), sodalite (30%) cancrinite (10%), plagioclase (5%), aegirina (2%), others (2%) 0.10
OdivGb Medium grained, homogeneous Ca-plagioclase (63%), Augite (25%), olivine (10%), others (2%). 0.06
AzB Vesicular texture Ca-plagioclase, pyroxene 8.90

Microcline from all granites is perthitic; the designation ‘‘others” includes altered minerals from feldspars and minor accessories such as sphene and iron oxides.
 Z.S.G. Silva, J.A.R. Simão / Construction and Building Materials 23 (2009) 3321–3327 3323

Table 2
Petrography and porosity of carbonate rocks.

Sample Texture Mineralogy Open porosity (%)

Ool ls Oolitic Calcite 3.05
Red ls Microcrystalline, bioclastic Calcite 0.24
Neg ls Microcrystalline, calci and bioclastic Calcite (94%), dolomite (6%) 1.18
Wm Crystalline, medium grained Calcite 0.17
Pkm Crystalline, fine to medium grained Calcite 0.20
Blkm Crystalline, fine Calcite 0.14

Fig. 1. Salt crystals deposited on the rock surfaces (magnification as indicated).

there on degradation is accelerated and rock disruption takes place, a fact that is After that it was possible to elaborate variation diagrams and to visualize the
partially responsible for the weight loss observed during the experiments. Fig. 3 changes and evolution of samples weight loss (in % per gram of sample) and com-
illustrates these features. pare the behaviour of different rock types shown in Fig. 4 and, for silicate and car-
bonate rocks, respectively.
2.6. Data on samples weight variation
2.6.1. Evolution and interpretation of the weight loss in igneous rocks
After each cycle of saline fog and after washing the samples to remove the salt Observation in Fig. 4 allows the following remarks: within rocks of granitic
excess, the rock samples were dried and weighed. The variations of weight recorded composition the porous Vila Real granites (respectively, SVRGr and AVRGr) were
throughout the experiment allowed an evaluation of the loss or gain of mass on dif- the ones which had greater mass loss during and at the end of the essay. AVRGr
ferent samples submitted to the salt spray. has registered almost always higher loss of weight relatively to the sound type
3324 Z.S.G. Silva, J.A.R. Simão / Construction and Building Materials 23 (2009) 3321–3327

Fig. 2. Salt crystals on the surfaces of nepheline syenite and odivelas gabbro (6) for all.

SVRGr. The former granite is an altered rock which was previously submitted to The rose granites, MonçGr and MonfGr, had a similar behaviour during the
hydrothermal and deuteric processes, leaving behind a weakened mineral struc- experiment concerning the loss of weight, as indicated by the parallel paths seen
ture. Biotites and iron oxides from the rock, once leached, impregnate the kaoli- on the diagram of Fig. 4. These rocks registered less loss of weight than the previous
nized feldspars, being responsible for the yellowish colour of the granite. granites. In fact, in the polished surface of these granites, there was almost no sign
Furthermore, the rock has a higher porosity which leads to an effective fluid circu- of alteration until the 9th cycle when the micas cleavage plans started opening.
lation and consequent salt crystallization and pore enlargement, in a gradual Foid syenites, though, had shown different aspects. The SoSy had a constant and
process. gradual loss of weight throughout the essay. This was essentially due to the lack of
At the 12th cycle both micas (biotite and muscovite) exhibit an expansion effect coherence of the grains of aegirine and biotite caused by the gradual crystallization
which in practice is registered through blistering of the cleavage plans and conse- of the salt in the cleavages and borders of the grains. The high amount of cancrinite
quent loss of material. This fact was also observed gradually in the AVRGr as the in the SoSy also contributed to the increase of alteration of the rock during the essay.
experiment proceeded. It should be noted, however, that after the 12th cycle, SVRGr The NeSy did not present alteration in the first cycles and registered an increase
experienced a sudden loss of weight exceeding that of AVRGr. Such fact is probably of mass from the 2nd to 4th cycle and from the 5th to 6th. After the 6th cycle the
due to the expansion and opening of the biotite and muscovite flakes, favoured by rock recovered the initial weight and, from there on starts losing weight at a high
the excellent cleavage (well observed with the aid of binocular microscope), as a re- rate, exceeding the SoSy from the 9th cycle on. The rock had an initial phase of
sult of the crystallization strength of the salt. deposition and crystallization of salt in the polished surface preferably along the
It is suggested that the differences could be accounted for the internal equilib- cleavage planes of orthoclase, aegirine–augite, biotites and in the fractures of neph-
rium in the two types of granites, in which the originally altered rock had a gradual eline. This feature was emphasized from the 9th cycle due to the disaggregation of
equilibrium process within mineral components and the sound granite, having a the mafic minerals (aegirine–augite and biotite), and was responsible for the visible
different rate of reaction between mineral phases and salt solutions was affected loss of weight of this rock in the final cycles of the experiment.
in a later stage. The basaltic rock (AzB) was the one that, among all igneous rocks tested, regis-
The CVGr registered about half of weight loss at the end of 13th cycle comparing tered higher loss of weight, similar to the AVRGr. This fact is possibly due to the par-
to other rock types. Originally this rock also presents hydrothermal alteration ticular structure and texture of the basalt, having a large exposed surface to the salt
developed as a late stage process that is reflected in the quality of the rock [15]. fog spray when compared to other igneous rocks. The rock has a vesicular texture,
The increasing of salt crystallization during the saline fog cycles occurred mainly and pore dimensions reach up to 5 mm of diameter. The force of crystallization of
in the cleavages planes of microcline and micas and in the twinning planes on the salt in the pores contributes to the disaggregation of the vitreous matrix mate-
plagioclases. rial and was responsible for its greater loss of weight.
 Z.S.G. Silva, J.A.R. Simão / Construction and Building Materials 23 (2009) 3321–3327 3325

Fig. 3. Rock surfaces after salt spray cycles (6) for all.

0.80

NeSy
0.70
OdivGb
MonçGr
0.60 AVRGr
CVGr
0.50 MonfGr
AzB
SoSy
0.40
BlGr
% SVRGr
0.30

0.20

0.10

0.00
1 2 3 4 5 6 7 8 9 10 11 12 13 14

-0.10

Cycles

Fig. 4. Evolution of the weight loss in igneous rocks (each cycle comprises 10 days each).

The loss of weight registered by BlGr was again gradual, similar to other gran- In contrast with silicate rocks just discussed, carbonate rocks clearly exhibit two
ites behaviour like the MonçGr, suggesting that the variation was essentially due to different trends of evolution regarding weight loss as response to salt spray expo-
the opening of cleavages of the biotites and hornblende, followed by the disruption sure. Marbles and compact, microcrystalline limestone, showing a progressive, very
of part of the grains of these minerals, as observed as the experiment proceeded. small and gradual change in weight and the oolitic and Negrais limestones, struc-
The OdivGb was the rock that registered the lowest weight loss among the turally very different from the previous ones, showing big differences in mass loss.
igneous rocks. This compact rock, with very low porosity, resisted to the saline The weight loss of marbles Wm, Pkm and Blkm and the limestone identified as
fog cycles almost without alteration on its surface and with relatively low loss of Red ls and the type of evolution displayed, similar during the 13 cycles is in agree-
mass. ment with the maintenance of the rocks surface features during the experiments. In
fact, this limestone being a compact rock, in terms of structure behaves very close to
2.6.2. Evolution of the loss of weight in the carbonate rocks marbles. The low porosity of all these rocks accounts for the low access of fluids
Diagram in Fig. 5 shows the evolution of the weight loss in the carbonate rocks available in the atmosphere which is a major factor of alteration as infiltration of
during the experiments. chlorides is the first step on this process of rock degradation.
3326 Z.S.G. Silva, J.A.R. Simão / Construction and Building Materials 23 (2009) 3321–3327

6.0

Wm

5.0 Pkm
Blkm
Ool ls
4.0 Red ls
Negr ls

% 3.0

2.0

1.0

0.0
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Cycles

Fig. 5. Evolution of the weight loss in carbonate rocks (each cycle comprises 10 days each).

Concerning the other limestones it is noteworthy the high loss of mass (about the rock after the salt cycles were completed were tried to detect
5.5% in the end of the experiment) of the Ool ls. In this case, again, the high porosity
mineral or rock structure changes, but they were inconclusive.
of rock allied to the easy access of fluid and crystallization of salt between micro-
fossils and oolites account for such a high loss of mass. Back weathering of the rock
The evolution of rock disaggregation during the salt spray
surface, rounding of edges and corners of the samples, outburst and rock fissures experiments showed that the limestones reaction was more effec-
were all features raised up after the 7th cycle. The other limestone variety, the Negr tive than that of silicates rocks and marbles and this is interpreted
ls registered about 3% of loss of mass. This loss is mostly related to the detachment as a function of the internal structure of the rocks. The stress from
of the clay fraction present in this rock, plus the enlargement of the cracks which
crystallization of salt within the pores was stronger in the lime-
are part of the internal characteristic of this limestone. This is a strong disruption
feature which starts early in the desegregation process, as the clayish material de- stones in which both types, the oolitic and the Negrais variety,
taches from the rock surface since the 7th and 8th cycles, promoting loss of the pol- have a particular pore network, favorable to fluid circulation. The
ished luster and uniform back weathering. pore size in these cases is a parameter to be taken into consider-
ation [18], followed by the capillary rise and evaporation effect,
3. Discussion and final remarks the cyclic wet and dry stages.
As a final statement it should be reminded that the mass varia-
Chloride penetration in silicate and carbonate rocks has shown tion which was detected and followed during the experiments dis-
to be a process similar to that occurring in concrete, in which case cussed in this paper is to be interpreted as a start point to consider
the development of microcracks due to the presence of chlorides when the rock behaviour has to be known in order to define con-
has been studied. In the case of concrete, two types of chlorides strains and proper uses of the rock. Microstructures and mechani-
are present, the so-called bound and free chlorides [16], which cal parameters can be highly influenced by the susceptibility of
act in different ways, either through fixing by adsorption or by porous building stones to salt crystallization as discussed and ap-
chemical reactions with the concrete components (the bound plied to some rock constructions [19].
type). The free chlorides, as the name indicates, can migrate and
are responsible for some type of corrosion. Regarding rock behav-
References
iour as related to that process, it is possible that both types of chlo-
rides are present and harmful to the rock. In the former case [1] Cardell C, Delalieux F, Roumpopoulos K, Auger F, Van Grieken R. Salt-induced
(bound types) because internal cracking will increase and pro- decay in calcareous stone monuments and buildings in a marine environment
motes alteration of the mineral components and leave the rock in SW France. Construct Build Mater 2003;17:165–79.
[2] Zezza F, Marcri F. Marine aerosol and stone decay. Sci Total Environ
fragile. On the latter (free type) because it will amplify any effect 1995;167:123–43.
on rock internal structure, depending on the internal pore distribu- [3] Lewin SZ. The mechanism of masonry decay through crystallization. In: Barkin
tion and on the mineralogy it will encounter along the migrating SM, editor. Conservation of historic stone buildings and
monuments. Washington, DC: National Academy of Sciences; 1981. p. 120–44.
path. [4] Mateus R, Silva ZCG, Fonseca M, Luís H, Cruz J, Jesus AP, et al. Técnicas
When applying this concept to the specimens studied it is Nucleares Aplicadas na Análise de Elementos Leves em Rochas Ornamentais.
understood that silicate rocks will be differently affected by salt In: Conferência Nacional da Física, Coimbra; 2006.
[5] Silva Z, Simão J. Syenite as ornamental stone – quality and alteration. In: Moore
infiltration as compared to the carbonate rocks. In carbonate rocks DP, Hungr O, editors. Engineering geology, a global view from the Pacific Rim,
the internal structure will be a strong parameter (since the miner- Proceedings of the 8th international congress of the international association
alogy is usually very simple), while in the case of silicate rocks be- for engineering and the environment, vol. IV, Vancouver: Balkema; 1998. p.
2897–901.
sides the rock fabric [17], the mineralogy will also be an important
[6] Simão J, Silva Z. Anorthosite and its use as dimension stone. Alterations as
factor to be taken into account. As an additional remark, minerals of response to weathering in polluted environment. In: Marinos PG, Koukis GC,
different composition will react in different ways, be they silica sat- Tsiambaos GC, Stournaras GC, editors. Engineering geology and environment.
urated or not, and the presence of minor elements in mineral struc- Proc. intern. symp., Balkema; 1997. p. 3257–62.
[7] Simão J, Silva Z. Composição Mineral e Alteração em Rochas Silicatadas; O
tures, combined with mineral framework may also dictate different ‘‘Granito Negro de Angola”. 1° Congresso Internacional da Pedra Natural. FIL,
paths and types of rock alteration. Thin section observations across Lisboa, Portugal, Rochas e Equipamentos, 1995, n. 43, p. 60–71.
 Z.S.G. Silva, J.A.R. Simão / Construction and Building Materials 23 (2009) 3321–3327 3327

[8] Gomes CSF, Silva JBP. Pedra Natural do Arquipélago da Madeira. Câmara de monuments en granite (Bretagne). In: Rodrigues D, Henriques F, Telmo
Lobos; 1997. Jeremias F. Proceedings, 7th international congress on deterioration and
[9] Simão JRS. Rochas Ígneas como pedra ornamental – Causas, condicionantes e conservation of stone, vol. 1; 1992. p. 129–38.
mecanismos de alteração. Implicações Tecnológicas. Tese de Doutoramento. [15] Silva Z, Simão J, Moreira A. Structural damage in building stone due to internal
Universidade Nova de Lisboa; 2003. mineral alteration. Castelo de Vide granite (Portugal) – a case study. In: 9th
[10] Azevedo HCA, Costa PHO. Catálogo de Rochas Ornamentais da Bahia- Congress of engineering geology for developing countries. Durban, South
Brasil. Bahia: Salvador; 1994. Africa: International Association for Engineering Geology and the
[11] Moura AC, Martins OR. Catálogo de Rochas Ornamentais Environment; 2002.
Portuguesas. Portugal: Ministério da Indústria e Energia; 1992. [16] Konin A, François R, Arliguie G. Penetration of chlorides in relation to the
[12] Arnold A, Kueng A. Crystallization and habits of salt efflorescences on wall, microcracking state into reinforced ordinary and high strength concrete. Mater
part I. Methods of investigation and habits. In: Proceedings of the 5th Struct 1998;31(June):310–6.
international congress on deterioration and conservation of stone; 1985. p. [17] Benavente D, Martinez-Martinez J, Cueto N, Garcia-del-Cura MA. Salt
255–67. weathering in dual-porosity building dolostones. Eng Geol 2007;94:215–26.
[13] Arnold A, Zehnder K. Crystallization and habits of salt efflorescences on wall, [18] Scherer GW. Stress from crystallization of salt. Cem Concr Res
part II. Methods of investigation and habits. In: Proceedings of the 5th 2004;34(9):1613–24.
international congress on deterioration and conservation of stone; 1985. p. [19] Theoulakis P, Moropoulou A. Microstructural and mechanical parameters
269–77. determining the susceptibility of porous building stones to salt decay.
[14] Robert M, Verges-Belmin V, Jaunet AM, Hervio M, Bromblet PH. Mise en Construct Build Mater 1997;11(1):65–71.
evidence de deux microssystème d’altération (chimique et physique) dans les