Content uploaded by Thirumalini Selvaraj
Author content
All content in this area was uploaded by Thirumalini Selvaraj on Aug 30, 2017
Content may be subject to copyright.
Knowing from the past –Ingredients and technology of ancient mortar
used in Vadakumnathan temple, Tirussur, Kerala, India
S. Thirumalini
a,1
,R.Ravi
b,2
, S.K. Sekar
c,
n
, M. Nambirajan
d,3
a
S.A. Engineering College, Chennai, Tamil Nadu, India
b
SRM University, Kattankulathur, India
c
School of Mechanical and Building Science, VIT University, Near Katpadiroad, Vellore 632014, India
d
Archaeological Survey of India, Office of the Director General, Janpath, New Delhi, 110011, India
article info
Article history:
Received 8 November 2014
Received in revised form
2 September 2015
Accepted 4 September 2015
Available online 11 September 2015
Keywords:
Ancient samples
Lime mortar
Chemical analysis
XRD
SEM
FT-IR and TGA with DTA
abstract
In the present study, lime mortar samples from the restoration site of Vadakumnathan temple, Kerala,
India have been analyzed. Samples from three different locations of the temple such as ancient wall,
gopuram and arch have been taken. Traditional methods such as chemical analysis, acid loss analysis and
organic test were conducted on mortar samples. Modern instrumental techniques such as electronic
particle size distribution, X-ray Diffraction (XRD), scanning electron microscopy coupled with energy
dispersive X-ray spectroscopy (SEM/EDX), Thermo Gravimetric Analysis (TGA), Differential Thermal
Analysis (DTA) and Infrared Spectroscopy (FT-IR) were employed in the study. The binder used in the
mortar is calcium high with 30% of clay mineral. A binder to aggregate ratio in the range of 1:1.5–2.5 has
been established from acid loss analysis. Particle of the aggregate are mostly silt in nature, hence nominal
sand would have been grinded to reduce the particle size and to induce pozzolanic reaction. The pre-
sence of carbohydrate, protein and fats are identified by organic test that are in agreement with FT-IR
analysis and TGA.
Calcite, aragonite and calcium complexes of silicate and aluminates in form of hydro thermal product
namely gyrolite and okenite are present in wall and gopuram samples. The formation of hydrothermal
products confirms that the mortar was produced by hot lime technology. In TGA, the decomposition of
CaCO
3
to CO
2
between 600 and 770 °C reveals the transformation of calcite from complex forms of CSH
(gyrolite and okenite) and CAH. The presence of degraded products such as syngenite and gypsum in
arch sample shows that the lime mortar is in complete deterioration where as mortar remained in good
condition in gopuram and wall samples.Texture along with elemental analysis (EDX) confirms the results
of chemical analysis.
&2015 Elsevier Ltd. All rights reserved.
1. Introduction
The architectural monuments of India, represent the cultural
heritage of the nation. The built heritage not only brings out the
artistic and technical capabilities of craftsmen but also the as-
pirations of people around the country. Vadakumnathan temple
(Fig. 1) at Tirussur, Kerala is one of the largest Lord Shiva temple,
constructed nearly 1300 years back by Parasurama, the sixth in-
carnation of Lord Vishnu. The temple is famous for its Kerala style
architecture and the exceptional work of craftsmanship, on the
gopurams, constructed on all four sides of the temple. In India, the
ancient texts like Citrasutra of the Vishnudharmottara, Silparatna
[1,2] talks about the construction of walls using lime mortar. As
mentioned in the ancient scripts like Silparatna and Vishudhar-
mottara, the temple was constructed with laterite blocks bonded
with lime mortar which consists of lime, sand and various
organics.
Strength and durability of lime mortar could be enhanced by
the addition of vegetable extract. Traditionally, people in India
have the practice of adding the natural organic admixtures from
the extracts of locally available plants, unrefined sugar and dry
seeds.
Ancient buildings were constructed using local and naturally
available materials. Stone, lime, clay, timber, palm leaves were
some of the materials used for the construction of temples in
Kerala. Also, ancient mortars comprises of various binders (or mix
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/jobe
Journal of Building Engineering
http://dx.doi.org/10.1016/j.jobe.2015.09.004
2352-7102/&2015 Elsevier Ltd. All rights reserved.
n
Corresponding author.
E-mail addresses: p.thirumalini@yahoo.in (S. Thirumalini),
ravistrucl@yahoo.co.in (R. Ravi), sksekar2011@gmail.com,
sksekar@vit.ac.in (S.K. Sekar), nambiasi@gmail.com (M. Nambirajan).
1
Mobile: þ91 9444135437; residence: þ91 4427640148.
2
Mobile: þ91 9444484937; residence: þ91 4422603185.
3
Mobile: þ91 9999881696.
Journal of Building Engineering 4 (2015) 101–112
of them) and natural or artificial aggregate along with herbal ad-
mixtures to enhance the strength and durability of the structures
[3].
Regional plants that are rich in carbohydrate, protein and fat
are generally added to the lime mortar mix. In central Kerala,
different types of herbs, namely Oonjalvalli, Kulamavu, Kadukai,
Pananchikai and jaggery (Unrefined sugar) were used as ad-
mixture in the temple for restoration work. Kadukai and jaggery
were added as carbon source where as other plant extracts are rich
in protein and fats. These are added to improve various properties
like carbonation, plasticity of the mix and also to enhance the
durability of the hardened mortar [4]. These mortars also served as
a platform for mural painting.
Although few research reports were published about the role of
organic admixtures on mortars, a thorough knowledge of lime
mortar used in ancient structures is essential for the preparation of
repair mortar. Many researchers have been working on the char-
acterization of ancient mortar, adopting various tools and techni-
ques on the physico-chemical, mechanical and mineralogical as-
pects. As the ancient mortar contains variety of raw materials
along with organics, it is difficult to understand the characteristics,
production and application techniques [5]. Even environmental
parameters such as climate and humidity could affect the strength
and durability of the mortar. All these parameters are interrelated
system that leads to complex problems during the analysis of
historic mortars.
In view of conservation and preparation of new restoration
mortars, it is essential to investigate the technology involved in
the traditional method of preparing lime mortars and plasters. It is
important that the mineralogical, chemical, micro-structural and
elemental composition has to be studied in order to understand
the procedure to produce restoration mortar similar to the original
one. The present study identifies the raw materials (lime, sand and
organics) used for the preparation of restoration mortar based on
the results of the characterization of the ancient mortars and its
application technology.
2. Background and climate of Tirussur
In the ancient and medieval period, Tirussur was the cultural
capital of Kerala. It has contributed very much in promoting trade
relationship with other parts of the world. It lies in the south-west
of Kerala with the geographical coordinates of 10°32’north and
76°32’east longitude which reveal the proximity of the region to
the equator. So the city does not face a major variation in the
temperature. Also there is no drastic change in the climate of the
region during summer and winter. The city records an average
rainfall of 3500 mm with a relative humidity of 95 percent.
3. Preliminary information about the construction of the
temple
It is essential to collect information from the sthapathis (ar-
chitect masters), as there is no documentation or information
about the traditional practice of lime mortar preparation. But the
stahpathis have knowledge about religious manuals of Indian
Fig. 1. (a) South entrance of Vadakumnathan temple (b) Birds view of the temple (c) Rama shrine (d) Restoration work at Rama shrine.
S. Thirumalini et al. / Journal of Building Engineering 4 (2015) 101–112102
origin that talk about the construction of temples in detail. Hence,
semi-formal interview was conducted with the sthapathis about
the material and the methods involved in the construction of
temple.
4. Methods and materials
4.1. Sampling of mortar
Totally fifteen samples, five samples from each selected loca-
tions of the wall plasters, gopuram and arch of Rama shrine were
collected for the study. Gopuram is the tower roof acting as en-
closure to the deity. Samples from gopuram part (Fig. 2a) were
taken from the top of Sri Rama shrine which was not opened since
the construction, but the wall (Fig. 2b and c) and arch samples
(Fig. 3a)would have been affected by environmental stress over
the centuries. Altered (crushed or powdered) and non-altered
mortar samples were taken from both internal and external por-
tion of the temple. Hence, analyzing gopuram samples gives de-
tails about the original mortar used whereas wall and arch sam-
ples gives information about the altered and degraded (mortars in
deterioration) mortars.
Samples have been collected from reasonably higher portions
of the temple in order to avoid the samples that were affected due
to the capillary rise of water and salts such as sodium chloride and
sodium sulphate. Sufficient quantities of samples were analysed to
account for the heterogeneity of the samples. The analysis was
carried out in a significant quantity of samples. Decay patterns and
cracks were encountered in mortars at arch location as in Fig. 3b
and c. Hence, sufficient care was taken to get mortar sample ad-
jacent to the area affected by cracks and fissures.
4.2. Acid digestion analysis
The composition of raw materials of the ancient mortar sample
was determined as per Rilem TC-167-COM [6] by dissolving the
carbonated mortar sample in 10% of hydrochloric acid, leaving out
fine aggregate fraction. The chemical analysis was carried out
binder dissolved filtrate.
4.3. Chemical analysis
Samples were analysed for their chemical composition by
adopting the traditional method of testing. CaO and MgO were
quantified on titration with Ethylene Diamine Tetra Acetic acid
(EDTA), using eBT as indicator. The amounts of Al
2
O
3
and Fe
2
O
3
were determined by Atomic Absorption Spectroscopy (AAS).
Gravimetric method was used to find the amount of silica present.
The hydraulic index (HI) and cementation index (CI) [7,8] were
calculated as per Eqs. (1) and (2) respectively.
Hydraulicindex,HI%AlO%FeO%SiO
/%CaO %MgO 1
23 23 2
=( + + )
(+ ) ()
Cementation index, CI 1. 1%Al O 0. 7%Fe O 2. 8%
SiO / %CaO %MgO 2
23 23
2
=( + +
)( + ) ()
4.4. Particle size analysis
As the quantity of sample available is limited, the fine ag-
gregate fraction left from acid loss test was used for particle size
analysis by using electronic particle size analyzer (CILAS 1064-Li-
quid) [9].
Fig. 2. (a) Exposed gopuram plasters after removing top slab (b) Wall (c) Another face of wall.
S. Thirumalini et al. / Journal of Building Engineering 4 (2015) 101–112 103
4.5. Organic analysis of lime plasters
The ancient mortars were tested for the presence of organics
such as carbohydrate, fat and protein [10]. The presence of fat and
protein were analysed by crude fat method and kjeldhal method
respectively. The carbohydrate content is calculated from the
percentage mass of protein and fat.
4.5.1. Determination of protein
Proteins are polymers of amino acids. Kjeldahl method is
usually considered to be the standard method of determining
protein concentration in the form of nitrogen concentration. The
kjeldahl method can conveniently be divided into three steps:
digestion, neutralization and titration. The sample is first digested
in strong sulfuric acid in the presence of a catalyst, which helps in
the conversion of the amine nitrogen to ammonium ions. In
neutralization, the ammonium ions are then converted into am-
monia gas, when heated and distilled. The ammonia gas was led
into a trapping solution where it dissolves and becomes an am-
monium ion once again. Finally the amount of the ammonia that
has been trapped is determined by titration with a standard
solution. As the number of moles of ammonia is same as number
of nitrogen, the amount of crude protein is given by %
Nitrogen 6.25 [11].
4.5.2. Determination of fat
The determination of fat in mortars are determined by crude fat
conforming to IS 7874-1975. The powder sample (5 g) heated at
105 72°C for at least 2 h was extracted with petroleum ether in a
Soxhlet extractor. Extraction was done at a condensation rate of 5–
6 drops per second for 4 h initially, and then 2–3 drops
per seconds for 16 h. The extract was dried on a steam-bath for
30 min, cooled in a desiccator and weighed as M
1
. Alternate drying
and weighing were done at 30 min intervals until the difference
between two successive weighing was less than 1 mg and the
lowest mass was noted as M
2
The mass of dried sample is taken as
m. The percent of crude fat content in the plant sample was cal-
culated as follows: 100 (M
1
M
2
)/m. [10].
4.5.3. Calculation of carbohydrate
The percentage of carbohydrate present in herbs was calculated
using Eq. (3).
ABCDTotal carbohydrate percentage 100 3
=( –(+++) ()
Where Ais percentage by mass of moisture
Bis percentage by mass of total protein
Cis percentage by mass of fat
Dis percentage by mass of total ash.[10]
4.6. Analytical techniques
Analyses of ancient mortars were done using following in-
strumental techniques. X-Ray Diffraction (XRD) analysis of finely
ground samples was done using Bruker Desktop-Diffractometer
working with the Cu Ka radiation and interpretation by Bruker
DIFFRAC.SUITE EVA Software. It gives a qualitative result on the
possible presence of minerals in the ancient lime mortar samples.
Thermal Gravimetric Analysis (TGA) was performed to find the
hygroscopic and structural bound water content of lime mortar,
and it has also helped to determine any possible thermal decom-
position of mineral phases [11]. The test was performed using
alumina (Al
2
O
3
) cells at a heating rate of 20 °C per min in a flushed
air atmosphere with a temperature range of 30–1000 °C. The
weight loss of the samples is monitored as a function of
Fig. 3. (a) Arch portion of Rama shrine (b) Visual crack on the top of the arch (c) Expansive nature of cracks at arch.
S. Thirumalini et al. / Journal of Building Engineering 4 (2015) 101–112104
temperature. Hence, TGA could serve as the best tool to char-
acterize and evaluate historic mortars. The results find a solution
for producing a compatible restoration mortars. At temperature
lesser than 120 °C, the weight loss is due to the release of hygro-
scopic water, 120–400 °C is due to structural bound water, 400–
550 °C is due to decarbonation of MgCO
3
and 600–800 °C is due to
the decarbonation of CaCO
3
. Differential Thermal Analysis (DTA)
reveals thermal transformation of various components heated
under controlled conditions. The phase transfers are exothermic or
endothermic in nature. DTA was carried out on the unaltered
mortar samples to find the dehydration, oxidation and decom-
position [12].
Infrared Spectroscopy (FT-IR) obtained from Perkin-Elmer 1000
instrument was used to get qualitative information on some of the
characteristic compounds present in the mortar sample. It in-
cludes the determination of calcium hydroxide, magnesium hy-
droxide, carbonates, gypsum, etc. and for identifying the presence
of salts (nitrates, sulfates, oxalates, etc.) as well as organic com-
pounds. Infrared Spectroscopy (FT-IR analysis) also provides sup-
plementary information to XRD. The principal constituents like
calcite, magnesite, portlandite and silica containing mineral (forms
of CSH), gypsum, syngenite and organics were identified in the
mortars. SEM (FEI Quanta FEG 200) was used to determine the
surface morphology of the mortars, particularly cracks and pores
in the samples. SEM images revealed the mineral morphology of
hydrated phases. EDX was used to unveil the elemental composi-
tion of mortars [13].
5. Results and discussions
5.1. Outcome of interview with sthapathis
Two or more forms of hydraulic lime or air lime were mixed in
equal proportion and used for mortar preparation.
Cocktail of organic extracts were mixed into the dry lime mortar
mix prior to slaking, whereas in other parts of the world slaking
was done with water.
Also the unique method of grinding the lime mortar mix prior
to slaking was practiced in the temple construction.
5.2. Binder –aggregate ratio
Dissolution of the mortar samples in hydrochloric acid yielded
two fractions: the insoluble fraction formed by siliceous ag-
gregates, the soluble fraction from calcium content and clay mi-
nerals [6].
In Vadakumnathan temple mortar samples, the binder to ag-
gregate ratio (Table 1) is in the range of 1:1.73–1:2.4 by weight. In
general, binder to aggregate ratio ranging from 1:1 to 1:6 was used
in historic mortar across the country. As discussed by Vitruvius
[14] in chapter IV of the Ten Books of Architecture, the best pro-
portion adopted for hydraulic mortars is 1:3 (by volume) mortar
which facilitates the maximum carbonation. Gameiro [15] has
expressed that calcite content are more in ratio of 1:3 mortar
compared to 1:2 mix as the higher aggregate content in 1:3 mix
increases the diffusion of carbon di oxide in to the mortar.
Whereas in the binder rich mortar (1:2), the co existence of un-
reacted portlandite and calcite is possible. In confirmation to the
above said statement, the XRD of gopuram indicates the presence
of portlandite. Hence, the presence of free portlandite in the go-
puram mortars shows that the carbonation process is still
proceeds.
According to Maravelaki et al. [16], ancient mortar in Crete and
Greece were constructed with a binder to aggregate ratio of 1:2–3
by weight. As discussed by Moropoulou [17], in a place without
drastic change in climatic conditions and high percentage of re-
lative humidity, the binder to aggregate ratio in the range of 1:1–
1: 3 is appropriate for good strength and durability. Also from the
visual examination and the test results, it is found that the mortar
is in good condition. Hence the mortar mix proportioning of
1:1.73–1:2.4 at Vadakumnathan temple is appropriate to the cli-
mate of the region (relative humidity of 95%), and this could be
one of the reason for durability and existence of historical struc-
tures over the centuries.
5.3. Chemical analysis
Representative samples from each category have been char-
acterized and presented in the Table 2. The major compounds
identified in the samples are CaCO
3
and SiO
2.
Presence of clay
minerals and impurities (Fe
2
O
3
and Al
2
O
3
) and SiO
2
around 30% in
the mortar samples assures the hydraulic character of the lime.
The silica content is found to be maximum, whereas magnesium
content is minimum. Further, the hydraulic and cementitious in-
dices of all the ancient samples are on the higher side and ranging
from 1.39 to 1.7 and 3.6 to 6 respectively. Higher the indices, the
lime will have more hydraulic property. This proves that the lime
putty used would be eminently hydraulic in nature similar to ce-
ment [8].
The ancient mortar under investigation contains calcite (CaCO
3
)
and a small amount of dolomite MgCa(CO
3
)
2
. Limestone contain-
ing MgCO
3
more than 20% of their weight is classified as dolomitic
lime while limestone with less than 5% MgCO
3
are categorized as
high calcium lime [18]. However, in all the samples, the magne-
sium oxide content is less than 2% which indicates the minimum
formation of dolomite. Hence, high calcium limestone was used as
a binder in the temple. However presence of small quantity of
magnesium oxide (MgO) would have improved the fresh state
properties of mortar such as plasticity and water retentivity which
contributes to the durability of the structure [18]. This also proves
that the ancestors have a thorough knowledge of the various
binder materials and their significance.
5.4. Particle size analysis
The particle size analysis shows similarities among the mortar
samples of the temple (Fig. 4).The grain size distribution shows
that most of the particles (75%) are in the range of 0.003–
0.064 mm (silt), whereas sand and clay particles accounts to
around 20% and 5% respectively. As the normal river sand used for
construction is coarser than silt, grinding the lime mortar mix
would have been carried out.
5.5. Organic analysis of mortar
High strength mortars (Hydraulic in nature) are preferred for a
place with high humidity [19], like Tirussur. Hydraulic lime is most
suitable for such a climate as it provides favorable condition for
the development of hydration products. Also the drawbacks in
Table 1
Comparative results of acid loss analysis of lime mortar samples.
Sample Initial Acid Wt after Wt of sand Wt of B/A
wt loss acid loss retained (A) binder (B) ratio
Wall 30.0 2.1 27.9 19.7 8.2 1:2.4
Gopuram 23.7 2.7 21.0 14.0 7.0 1:2
Arch 30.0 1.3 28.7 18.2 10.5 1:1.73
B/A –Binder to aggregate ratio.
S. Thirumalini et al. / Journal of Building Engineering 4 (2015) 101–112 105
using lime mortar such as slow setting, low workability, lesser
compressive strength and low carbonation could be counteracted
by the addition of natural organic admixtures that are regionally
available [20]. The organic test (Table 3) results shows the pre-
sence of carbohydrates, protein in all the mortar samples. Also, the
loss of ignition (LOI) at 400 °C which accounts for organics is
around 2.5% in the ancient mortars (Table 2). Hence different plant
extracts would have been added as an admixture to improve
workability, early setting time, enhance carbonation and durability
of the mortar.
In all the ancient mortar samples, protein and fat contents are
more or less similar. But the carbohydrate content in gopuram
samples are higher compared to wall and arch samples, which
could be due to the fact that gopuram is not exposed to the at-
mospheric CO
2
, resulting in incomplete carbonation. As a con-
sequence, the carbohydrate content of the organics added into the
lime mortar mix has been consumed in lesser quantities in go-
puram. Still in Central Kerala, there is a traditional practice of using
fermented cocktail of plant extracts which are rich in carbohy-
drates, protein and fats in the restoration works of monuments.
The locally available herbs namely Oonjalvalli (Cissus glauca Roxb),
Pananchikaai (Cochlospermum religiosum), Kulamavu (Perseama-
crantha), Kadukai Terminalia chebula and palm jaggery from palm
tree would have been added along with lime mortar as bio-ad-
mixtures to improve its functional properties. The fermentation of
herbs would have been carried out for 7–15 days. It reduces the
carbohydrates present in herbs to carbon-di-oxide and thereby
enhances the carbonation within the mortar [21]. The carbohy-
drates are magnificent adhesives and a natural super plasticizer.
They react with Ca(OH)
2
of lime and forms network of gel that has
resulted in better binding of the solid microstructure. Carbohy-
drates have moisturizing capacity that creates an environment,
which not only enhances the hydraulic component formation, but
also controls the shrinkage, thus avoiding micro cracking within
the structure [22].
Jozefjasiczak and Krzysztof Zielinski [23] have stated that the
addition of proteins acts as air entraining agent in fresh mortars
and increases the workability. Fat acts as water proof to mortar
and controls the water movement. Proteins improve the hydro-
phobic property of the mortar [22]. Also, the calcium complexes
formed during the interaction of proteins with the divalent cal-
cium ions contribute to the reduced water absorption, in the same
way as in the case of synthetic polymers. The proteinaceous ma-
terial present in the lime mortar samples will convert calcium
oxide into calcium oxalate. Proteins react chemically with clay by
exchanging inorganic cations of the clay with organic, resulting in
a mechanism relating to the ability of amino acids (amides) to
encourage clay flocculation [9]. However, the porosity of the
mortar is not much affected. Hence, the plant extracts protects the
temple structure from environmental deterioration.
5.6. Analytical techniques
From the XRD results (Fig. 5) of wall samples, presence of cal-
cite, calcium silicate hydrates (CSH) such as gyrolite and okenite,
Table 2
Chemical composition of various minerals present in lime mortar.
Sample CaO MgO Al
2
O
3
Fe
2
O
3
SiO
2
CO
2
CaCO
3
LOI HI CI
Wall 28.33 0.22 1.28 2.22 36.07 11.95 50.43 2.07 1.39 3.6
Gopuram 21.26 0.36 0.77 1.62 45.5 22.81 37.95 2.04 2.21 5.9
Arch 23.70 0.27 1.16 1.68 39.80 25.28 42.30 2.69 1.78 4.8
CaOþMgO: Binder, Al
2
O
3
þFe
2
O
3
þSiO
2
: Clay Minerals, LOI: Loss on Ignition at 400 °C, HI: Hydraulic Index, CI: Cementation Index.
100
80
% Finer
60
40
20
0
1.0 10.0 100.0
0.1
0.04
Diameter(µm)
500.0
0
Fig. 4. Particle size distribution of wall, gopuram and arch.
Table 3
Presence of organics in historic mortar.
Description Carbohydrates (%) Protein (%) Fats (%)
Wall 10.8 5.55 2.3
Gopuram 14.6 4.8 3.2
Arch 9.2 5.3 2.7
S. Thirumalini et al. / Journal of Building Engineering 4 (2015) 101–112106
calcium aluminum hydrates (CAH) have been identified. Two main
peaks are observed, one at 290 °C for CSH and CAH and that of
calcite transformation in the range of 745–765 °C. The CSH gel can
be stable in natural form of gyrolite and okenite [24]. The forma-
tion of hydrothermal products (gyrolite and okenite) is possible
when lime is manufactured by hot lime technology [25]. Hence,
the high intense peaks of CSH and CAH observed in XRD also in-
dicates the pozzolanic reaction due to addition of shell lime during
burning, followed by hot lime technology and subsequent grinding
of the mortar mix.
Mix of two forms of lime (limestone and shell lime), sand and
organic extract would have been slaked for a longer duration, due
to which the lime and sand particles are broken down from micro
to nano scale [26]. This is because the temperature rises up to
180 °C during slaking and lot of heat has been generated that
causes the lime to split resulting in the formation of hydrothermal
products of CSH. This process is termed as hot lime technology.
This process will be incomplete in case of improper slaking and
hence to enhance the lime silica interaction, grinding would have
been carried out at the time of placing the mortar. In support to
the above statement, particle size analysis indicates the presence
of more amount of finer particles in the mortar sample resulting
from grinding process.
By grinding, splitting of lime and silica grains takes places and
become finer in order of nanometers. Fine silica and clay minerals
accelerate the reaction rate less than a day to form CSH phases
without changing the microstructure of lime matrix [26]. Without
grinding, the mortars will take much time to reach the similar
extent of reaction. Similarly, micronized lime will result in change
in shape of portlandite for good workability [17].
In all the Vadakumnathan mortar samples (Table 4), the weight
loss is around a reaction temp of 750 °C (600–800 °C) showing
decomposition of CaCO
3
and release of CO
2
. In the case of air lime
mortar, there will be direct reaction of CaO with CO
2
to form pure
CaCO
3
. The thermal break down of calcite and subsequent release
of CO
2
will show distinctive endothermic peak around 840 °C. The
reason for early decomposition is that the formation CaCO
3
along
with hydraulic components like CSH and CAH are formed at early
ages. As the age processes the complex forms of CAH and CSH
phases re-carbonate and forms CaCO
3
. Early decomposition of
CaCO
3
to CO
2
between 600 °C and 770 °C and the major weight
loss in this range shows that the hydraulic components (CSH like
gyrolite and okenite and CAH) are completely transformed into
CaCO
3
.
It also reveals complete transformation of calcite from complex
forms of CSH. TGA data presented in Fig.6 (wall mortar) indicates
the major weight loss in the range of 600–755 °C and it is due to
decomposition of calcite and release of CO
2
. When heated up to
120 °C, the mortar containing less than or equal to 2% of absorbed
water demonstrate the absence of hygroscopic character, indicat-
ing lesser amount of portlandite in the wall and arch compared to
gopuram mortars.
Presence of portlandite (Ca(OH)
2
) in gopuram samples detected
at a dehydration peak of 388 °C in TGA is validated by the results of
XRD and SEM. Also, the FT-IR vibrations present at 1011 cm
1
is
due to hydraulic compounds detected only in gopuram and not in
wall and arch samples. As the lime mortar samples are taken from
the inner part, the possibility of deteriorating agents like wind,
dust and dirt, rain water and atmospheric gases affecting the
mortar are less significant. The samples taken at gopuram are not
directly exposed to atmospheric air and the humid environment.
Hence carbonation is delayed, and portlandite crystals are ob-
served (Fig. 7). But, in the gopuram samples the complete trans-
formation to calcite has not occurred. Presence of portlandite and
calcium carbonate together helps in self-healing of mortar that
enhances the durability [27].
However, in arch samples, peaks of XRD images indicate the
formation of degradation products such as syngenite
(K
2
Ca(SO
4
)
2
.2H
2
O), gypsum (CaSO
4
) along with major calcite peaks
throughout the range. Presence of calcium aluminate silicate hy-
drate (CASH) and CSH is detected in the form of vesuvianite and
overite respectively. Long term exposure of the temple to atmo-
spheric pollutants has resulted in the formation of degradation
products. Absorption bands at 714, 874 and 1420 cm
1
indicates
the presence of calcite. The wavelength at 970–100 0 cm
1
arises
from CSH vibrations and absorption band attributed to gypsum are
observed at 3548, 3420, 1143, 1117, 1017, 670 and 603 cm
1
.The
wave bands at 1624 and 713 cm
1
may be of amides (proteins)
and calcite respectively. Other bands at 3248, 753 and 643 cm
1
Fig.5. XRD of ancient lime mortar samples A-arch, G-gopuram and W-wall. GYR –
Gyrolite, O –Okenite, C –Calcite , CMC –Calcium/Magnesium Carbonate, P –
Portlandite CAH –Calcium Alumium Hydrate , A –Aragonite S –Syngenite and GP –
Gypsum.
Table 4
Results of thermal analysis.
Sample Weight loss per temperature range (%)
o120 °C120–400 °C400–550 °C600–800 °C
Wall 1.79 2.11 2.67 91.99
Gopuram 5.9 10.80 7.95 72.04
Arch 6.78 15.07 6.99 65.43
Fig. 6. TGA and DTA graph of wall samples.
S. Thirumalini et al. / Journal of Building Engineering 4 (2015) 101–112 107
indicate the presence of degrading component, syngenite [28].In
TGA (Fig. 8), mortar analysis shows the characteristic peak of
gypsum dehydration at 119–139 °C followed by peak of CaCO
3
,
decomposition at 728 °C which is in confirmation with XRD. The
peaks about 150 °C is due to overlapping of CSH and ettringite
minerals. Endothermic phenomena at 290–400 °C are mainly due
to the thermal decomposition of the syngenite which is possible
only in hydraulic mortars. It also indicates the formation of a new
complex of calcium aluminum silicate hydrate called Vesuvianite.
Long fibrils of gypsum [29] and syngenite crystals are observed in
SEM images (Fig. 9).
The XRD results identify the formation of deterioration mi-
nerals such as syngenite and gypsum in the arch portion indicating
that the mortar is not structurally sound. The EDX results of arch
samples also validate the results of XRD as it has the elemental
composition of sulphur and potassium which is not found in other
samples (gopuram and wall). The formation of gypsum and syn-
genite leads to expansive forces during its growth causing ex-
tensive cracks and spalls as seen in photographs of arch portion of
temple (Fig. 3b and c). These results are being further reinforced
with the SEM images (Fig. 9) of arch samples showing long fibrils
of gypsum [29].
From the DTA graph, the dehydration range of alumino silicates
and clay minerals from 200 °C to 650 °C confirms the pozzolanic
reaction. Finely ground shell lime available in coastal areas of
Kerala containing clay impurities of magnesium alumino silicates
would have been mixed and burnt with limes to initiate the
pozzolanic action. Also the presence of potassium in small quan-
tity as indicated by EDX results ensures that the clay or shell lime
is mixed with limestone to enhance the pozzolanic reaction [30].
The evolution of lime mortar starts from burning followed by
wet slaking (hot lime technology), and subsequent grinding. Nat-
ural hydraulic lime is made from lime stone that naturally contains
clay mineral (Si
x
Fe
y
Al
z
). Natural hydraulic lime (NHL) gains its
initial strength by hydraulic set followed by strength gain under
carbonation. Momade and Atiemo [31] claims that during lime
burning process, structure of clay particles are broken down at a
temperature 500–600 °C, making the substances chemically re-
active. During this process, the hydroxyl ions escapes out, there by
distorting the crystal structure of the clay, finally converting into
active pozzolan. The active clay imparts early strength to the
binder.
Hence in confirmation with XRD, the TGA results indicates the
shell lime rich in clay impurities along with limestone dehydrox-
ylate during burning process and acts as a pozzolan imparting
early strength to the mortar. Also, the dehydration of magnesia
between 250 °C and 550 °C is seen in all the mortar samples that
substantially contribute to desirable plasticity [32].
The FT-IR results of ancient samples are given in Fig. 10. In FT-IR
spectrum, the intense peaks for calcite at 711.73, 873.75, 1436.97,
1795.73 and 2513.25 cm
1
are found in the wall sample. Calcite is
obtained from carbonated lime, silica and feldspar from ag-
gregates. Silica containing minerals are found at 468 cm
1
.
In support to organic analysis (Table 3), the peaks 1631.78, 2987
and 2873.94 cm
1
in FT-IR corresponds to organics. The
1631 cm
1
peak indicates the presence of amides (proteins)
whereas 2987 and 2873 cm
1
indicates carbohydrates or poly-
saccharides. In confirmation with organic test and FT-IR, the exo-
thermic effect within the of range 300–400 °C in the TGA (Fig.11)
of gopuram indicates the presence of the organic matter.
Clear evidence of a lime silica matrix is presented in Scanning
Electron Micrographs (Fig. 12) showing the typical hydraulic
compounds (CSH and CAH). Amorphous calcium carbonate is
Fig. 7. SEM images of unaltered gopuram mortars (a) and (b) aragonite (Calcite) fibrous crystals with portlandite.
Fig. 8. TGA and DTA images of degraded mortar at arch.
S. Thirumalini et al. / Journal of Building Engineering 4 (2015) 101–112108
Fig. 9. SEM Images of degraded mortar at arch (a) and (b) Long fibrils of gypsum and syngenite crystals (c) Vesuvanite (CASH) (d) formation of small plate like aggregates of
calcium silica gels.
Fig. 10. FT-IR of historic samples: A- Arch, G- gopuram and W-wall.
S. Thirumalini et al. / Journal of Building Engineering 4 (2015) 101–112 109
found with some perfect crystals of the calcite. It is found that the
co-existence of amorphous and crystalline calcite enhances the
hydration process as more amounts of free lime has been available
for reaction [33].
The EDX results of wall (W
1
–W
6
), gopuram (G
1
–G
4
) and arch
(A
1
–A
8
) samples are presented in Table 5. Calcium, magnesium,
silica, alumina, ferrous and oxygen are present in gopuram and
wall samples. In arch samples, along with above said elements,
Fig. 11. TGA and DTA images of unaltered gopuram mortars.
Fig. 12. SEM of wall samples (a) cementitious lime silica matrix (b) white heap of calcite on sand grain (c) possible gyrolite crystal.
S. Thirumalini et al. / Journal of Building Engineering 4 (2015) 101–112110
sulfur and potassium were also present. Hence in gopuram and
wall samples, the formation of CSH, CAH and calcite is possible
whereas in arch, ettringite and syngenite also exists. Also from the
visual examination, the gopuram mortar is found to be original
and remains in good condition as it is encapsulated inside a
structure (not exposed to atmosphere).
Further, the results substantiate the formation of CSH and CAH
minerals as interpreted in XRD, TGA and SEM.
Though EDX can detect the elements having atomic mass
greater than 7, the results indicates higher amount of carbon up to
26%. The value can be justified as the formation of calcite requires
12 no of carbon, and the remaining may contributed to presence of
organics used in mortar. As already discussed, the presence of
organic admixtures enhances the formation of CSH, and CAH.
Hydrates of calcium silicates and calcium aluminates could modify
the texture and microstructure of lime mortar mix, resulting in
increased hydraulic properties, ultimately resulting in enhanced
mechanical strength. Moreover, CAH imparts compactness and
high flexibility to the mix. Addition of carbohydrates also retard
the formation of ettringite (which would lead to the formation of
cracks) and enhance the formation of calcium silicate hydrate
(CSH) and calcium aluminum hydrate (CAH) [33,26].
6. Conclusion
The ancient mortar of Vadakumnathan temple has binder to
aggregate ratio of 1:1.73–2.4. Presence of clay minerals around 30%
in the binder indicates its eminently hydraulic nature. Finely
ground shell lime rich in clay mineral could have been mixed and
burnt with limestone to initiate the pozzolanic action and also
impart hydraulic character to the mortars. The particle size ana-
lysis of the ancient mortar reveals the presence of silt particles
(more fine particles) around 75% which could be due to the slaking
of shell lime, with lime stone along with silica aggregates and also
attributed by the grinding of the mortar mixture. Organics in the
form of carbohydrates, proteins and fats have been identified in
the mortars. Carbohydrates on fermentation enhances carbonation
within the mortar whereas proteins acts as air entraining agent in
fresh mortars and increases the workability and fat acts as water
proof to mortar and controls the water movement. Organics has
played a vital role in formation of hydrated phases and also in
resisting the environmental degradation. The mortar is complex
multi-phase composites, comprising of abundant formation of
complex CASH and CSH in the form of gyrolite and okenite and
amorphous phase of calcite along with silica aggregates. The for-
mation of hydrothermal forms CSH and CAH in the mortar suggest
that the lime is produced by hot lime technology. Presence of both
portlandite and calcite in the mortar samples contributes to self-
healing of lime mortar which enhances the durability of the
mortar.
Wall and gopuram mortars of the temple are in good condition
showing complete carbonation and performing better in a rela-
tively high humidity and wet environment. The intensity of dif-
fraction peaks of the historic samples (wall and gopuram) suggests
that there is great amount of amorphous calcium carbonate in the
sample. But, arch locations of the temple show extensive cracks
and spalls. Presence of syngenite and gypsum in arch portion
shows that the mortar is in complete damaged stage, and it needs
attention. However, the main reason for survival of the temple
over centuries was the appropriate construction material used
with hot lime technology.
Acknowledgment
Our sincere thanks to the staff members of Archaeological
Survey of India, Tirussur circle for their cooperation and support to
collect the information and samples. The author also likes to thank
RAMCO Research and Development Centre and Ms. Karunya, for
help rendered in using XRD.
References
[1] Sivaramamurthi, Vishnudharmottara of Sixth century, 1978.
[2] Bhattacharya, Silparatna of sixteenth century, 1974.
[3] A. Moropoulou, A. Bakolas, S. Anagnostopoulou, Composite materials in an-
cient structures, Cement Concr. Compos. 27 (2005) 295–300.
[4] P. Thirumalini, R. Ravi, S.K. Sekar, M. Nambirajan, Study on performance en-
hancement of lime mortar used in ancient temples and monuments of India,
Indian J. Sci. Technol. 4 (2011) 1484–1487.
[5] Ioanna Papayiann, Vasiliki Pachta, Maria Stefanidou, Analysis of ancient
mortars and design of compatible repair mortars: the case study of Odeion of
the archaeological site of dion, Constr. Build. Mater. 40 (2013) 84–92.
[6] B. Middendorf, J.J. Hughes, K. Callebaut, G. Baronio, I. Papayiaani, Investi-
gation methods for the characterization of historic mortars –Part 2:
chemical characterisation, Rilem TC-167-COM, Mater. Struct. 38 (2005)
771–780.
[7] E. Cechova, The Effect of Linseed Oil on the Properties of Lime Based Re-
storation Mortars (Ph.D. thesis), University of Bologna, Italy, 2009.
Table 5
Possible elements in EDX in weight.
Sample C O Si Al Fe Ca Mg S K Ti
W
1
25.02 47.17 0.40 –– 27.4 –– – –
W
2
23.26 46.77 0.68 –– 29.29 –– –
W
3
22.63 41.69 0.84 –– 34.83 –– – –
W
4
22.68 46.00 0.41 –– 30.92 –– – –
W
5
19.12 47.54 29.79 –– 03.55 –– – –
W
6
26.06 43.96 03.01 –– 26.96- –– – –
G
1
25.45 47.01 0.52 –0.70 26.32 –– – –
G
2
11.24 52.44 30.25 –0.47 03.18 –– 2.43 –
G
3
18.06 38.45 2.00 –0.45 40.15 –– 2.51 –
G
4
14.83 45.97 31.29 –0.49 3.51 –– – –
A
1
10.00 32.44 0.34 –– 15.65 –18.77 22.80
A
2
11.84 30.62 0.67 –– 14.44 –18.94 23.48 –
A
3
14.22 40.88 2.06 1.87 0.89 13.15 –12.54 –
A
4
19.58 37.84 2.69 1.92 1.35 34.50 –1.06 01.06 –
A
5
15.47 27.49 4.47 3.40 9.04 28.95 1.31 1.48 2.11 2.48
A
6
15.47 39.49 2.81 1.79 1.32 36.76 - 1.03 1.32 –
A
7
17.50 27.90 4.20 3.08 6.13 33.11 0.92 1.13 1.74 1.00
A
8
18.91 30.21 7.18 6.68 12.16 12.73 1.56 0.97 1.75 3.89
C–Carbon O –Oxygen Si –Silica Al –Aluminum Fe –Iron Ca –Calcium Mg –Magnesium S ‐Sulphur K –Potassium Ti –Titanium.
S. Thirumalini et al. / Journal of Building Engineering 4 (2015) 101–112 111
[8] R.M. Lawrence, T.J. Mays, S.P. Rigby, P. Walker, D.D. Ayala, Effect of carbonation
on the pore structure of non hydraulic Lime mortars, Cement Concr. Res. 37
(2007) 1059–106 9.
[9] M. Singh, B.R. Arbad, Characterization of traditional mud mortar of the
decorated wall surfaces of Elora Caves, Constr. Build. Mater. 65 (2014)
384–395.
[10] IS:7874 (Part I)-1975, Methods of tests for Animal feeds and feeding stuffs,
Bureau of Indian Standard, New Delhi, India, 1991.
[11] G. Cultrone, E. Sebastain, M. Ortega, Forced and natural carbonation of lime
based mortars with and without additives: mineralogical and textural chan-
ges, Cement Concr. Res. 35 (2005) 2278–2289.
[12] A. Morpoulou, A. Bakolas, K. Bisbikou, Characterization of ancient byzantine
and later historic mortars by thermal and X ray diffraction techniques, Ther-
mochim. Acta 269/270 (1995) 779–795.
[13] W. Bartz, T. Filar, Mineralogical characterization of rendering mortars from
decorative details of a baroque building in Kozuchow (SW Poland), Mater.
Charact. 61 (2010) 105–115 .
[14] P. Vitruvius, Ten Books on Architecture Translated by Ingrid D Rowland,
Cambridge University press, United Kingdom, 1999.
[15] A. Gamerio, A. SantosSilva, P. FariaJ.Grilo, T. Branco, R. Veiga, A. Velosa, Physical
and chemical assessment of Lime –metakaolin mortars: Influence of binder:
aggregate ratio, Cement Concr. Compos. 45 (2014) 264–271.
[16] P. Maravelaki-Kalaitzaki, A. Bakolas, I. Karatasios, V. Kilikoglou, Hydraulic lime
mortars for the restoration of historic masonry in crete, Cement Concr. Res. 35
(2005) 1577–1586.
[17] A. Moropoulou, A. Bakolas, K. Bisbikou, Physico-chemical adhesion and co-
hesion bonds in joint mortars imparting durability to the historic structures,
Constr. Build. Mater. 14 (2000) 35–46.
[18] ElifUgurlu, HasanBoke, The use of brick-lime plasters and their relevance to
climatic conditions of historic bath buildings, Constr. Build. Mater. 23 (2009)
2442–2450.
[19] Siva Sankar Panda, Pradip Kumar Mohapatra, Rishabh Kumar Chaturvedi,
Sushanta Kumar Kar, Chemical analysis of ancient mortar form excavation
sites of kondapur, Andhra Pradesh, India to understand the technology and
ingredients, Curr. Sci. 105 (2013) 837–842.
[20] A. El-Turki, R.J. Ball, G.C. Allen, The influence of relative humidity on structural
and chemical changes during carbonation of hydraulic lime, Cement Concr.
Res. 37 (2007) 1233–1240.
[21] P. ThirumaliNI, S.K. SekAR, Effect of microbiologically induced CaCO
3
pre-
cipitationby addition of kadukai and jaggery in Lime mortar, Int. J. Appl. Eng.
Res. 9 (2014) 6149–6157.
[22] S. Chandra, J. Aavik, Influence of proteins on some properties of portlandce-
ment mortar, Int. J. Cement Compos. Lightweight Concr. 9 (1987) 91–94.
[23] Jozef Jasiczak, Krzysztof Zielinski, Effect of protein additive on properties of
mortars, Cement Concr. Compos. 28 (2006) 451–457.
[24] Georgia Zacharopoulou, Interperting chemistry and technology of lime binders
and implanting it in the conservation field, Conserv. Partrimonio 10 (2009) 37–48.
[25] Goreti Margalha, Rosario Veiga, Antonio Santos Silva, Jorge de Brito, Tradi-
tional methods of mortar preparation: the hot lime mix method, Cement
Concr. Compos. 33 (2011) 796–804.
[26] A. Moropoulou, A. Cakmak, K.C. Labropoulos, R. VanGrieken, K. Torfs, Ac-
celerated microstructural evolution of a calcium–silicate–hydrate (C–S–H)
phase in pozzolanic pastes using fine siliceous sources: comparison with
historicpozzolanic mortars, Cement Concr. Res. 34 (2004) 1–6.
[27] B. Lubelli, T.G. Nijland, R.P.J. VanHees, Self healing of lime based mortars:
microscopy observations on case studies, HERON J. 56 (2011) 75–91.
[28] J. Lanas, R. Sirera, J.I. Alvarez, Compositional changes in lime based mortar
exposed to different environments, Thermochim. Acta 429 (2005) 219–226.
[29] Hasan Boke, Sedak Akkurt, Ettringite formation in historic bath brick –lime
plaster, Cement Concr. Res. 33 (2003) 1457–1464.
[30] Achia Khanom, D.K. Saha, S.M. Rahaman, A. Mamun, Elemental analysis of old
mortar used in various archaeological sites of Bangladesh by SEM and EDX, J.
Asia. Soc. Bangladesh Sci. 389 (2012) 145–153 .
[31] Z.G. Momade, E. Atiemo, Evaluation of pozzolanic activity of some clays in
Ghana, J. Sci. Technol. 24 (2004) 76–83.
[32] K. Luke, Phase studies of pozzolanic stabilized calcium silicate hydrate at
180 °C, Cement and Concr. Res. 34 (2004) 1725–1732 .
[33] Yuyao Zeng, Bingjian Zhang, Xiaolin Liang, A case study and mechanism in-
vestigation of typical mortars used on ancient architecture in China, Ther-
mochim. Acta 473 (2008) 1–6.
S. Thirumalini et al. / Journal of Building Engineering 4 (2015) 101–112112