T. IÞÝK, E. U. SAÐÝN: CHARACTERISTICS OF BYZANTINE-PERIOD LIME MORTARS ...
461–470
CHARACTERISTICS OF BYZANTINE-PERIOD LIME MORTARS
AND PLASTERS FROM THE ANAIA CHURCH (KADIKALESI)
ZNA^ILNOSTI APNENIH MALT IN OMETOV S CERKVE ANAIA
IZ BIZANTINSKEGA OBDOBJA
Tuðçe Iþýk
*
, Elif Uðurlu Saðýn
Ýzmir Institute of Technology, Faculty of Architecture, Department of Conservation and Restoration of Cultural Heritage, Urla, Izmir, Turkey
Prejem rokopisa – received: 2022-03-15; sprejem za objavo – accepted for publication: 2022-06-10
doi:10.17222/mit.2022.524
In this study, Byzantine-period lime mortars and plasters used in the Anaia Church in Kuºadasý-A yd ýn were examined in order
to determine their characteristics and investigate the continuity of the lime mortar technology through centuries in the Anaia
Church. The results will also contribute to future conservation studies at the site. Basic physical properties, raw-material compo-
sitions and hydraulic properties of lime mortars and plasters; mineralogical and chemical compositions, microstructural proper-
ties of binders, aggregates and limes; and pozzolanic activities of aggregates were determined using RILEM test methods, XRD,
SEM-EDS and TGA. Mortar samples were comprised of natural aggregates whereas lime plasters were made of brick aggre-
gates. Analyses revealed that plasters were slightly less dense and more porous than mortars due to the porous structure of the
brick aggregates. All mortars and plasters were hydraulic due to the use of highly reactive pozzolanic aggregates. The basic
physical properties, raw-material compositions, mineralogical and chemical compositions of mortars and plasters were found to
be similar throughout the construction periods spread over different centuries. These similarities revealed the conscious knowl-
edge of the lime mortar technology during the Byzantine period in Western Anatolia.
Keywords: Byzantine period, hydraulic lime mortar, pozzolan, characterization
Predstavljena {tudija preu~uje apnene malte in omete s cerkve Anaia v kraju Kuºadasý-A yd ýn, ki so iz bizantinskega obdobja.
Namen {tudije je dolo~itev lastnosti the materialov in preu~evanje tehnologije apnenih malt na cerkvi skozi stoletja. Izsledki
{tudije bodo prispevali h konservatorskim pristopom na cerkvi v bodo~e. Osnovne fizikalne lastnosti, surovinsko sestavo in
hidravli~ne lastnosti apnenih malt in ometov; mineralo{ko in kemijsko sestavo, mikrostrukturne lastnosti veziv, agregatov in vrst
apna ter pucolanske aktivnosti agregatov; smo dolo~ili z RILEM-ovi testnimi metodami, XRD, SEM-EDS in TGA. Maltni
vzorci so vsebovali naravne agregate, ometi pa so bili pripravljeni z ope~nimi agregati. Z analizami smo ugotovili, da imajo
ometi manj{o gostoto in so bolj porozni kot malte. To lahko pripi{emo porozni strukturi ope~nih agregatov. Vsi ometi in malte
vsebujejo hidravli~ne produkte, zaradi visoko reaktivnih pucolanskih agregatov. Osnovne fizikalne lastnosti, sestave osnovnih
materialov, mineralo{ke in kemijske sestave malt in ometov se skozi stoletja niso spreminjale, kar nakazuje na ohranjanje znanj
o tehnologiji apnenih malt med bizantinskim obdobjem v Zahodni Anatoliji.
Klju~ne besede: bizantinsko obdobje, hidravli~na apnena malta, pucolan, karakterizacija
1 INTRODUCTION
Historical lime mortars and plasters were produced
using non-hydraulic lime and natural or artificial
pozzolanic aggregates. Amorphous silica and alumina in
the structures of pozzolans react with lime in the pres-
ence of water, and calcium silicate hydrate (CSH) and
calcium aluminate hydrates (CAH) that provide hydrau-
lic properties to mortars and plasters are formed as the
result of this reaction.
1,2
Hydraulic, mechanical, and
microstructural properties of lime mortars produced us-
ing pozzolanic aggregates had been appreciated by dif-
ferent civilizations like Romans, Byzantines and Otto-
mans. Lime mortars provided stability and durability
especially to water-related structures throughout centu-
ries until the invention of modern cement.
3
It is known that lime had been produced in the kilns
situated close to the raw material sources, and lime mor-
tars had been manufactured close to the building sites by
local craftsmen.
4
Therefore, they can be accepted as the
characteristics of the region and the period, in which
they were produced. Characterization of lime mortars
and plasters and determination of the chemical, mineral-
ogical and microstructural properties of their raw materi-
als will provide historical information about their pro-
duction as well as a better understanding of the
structures applied.
This study aims to determine the physical, mineral-
ogical and hydraulic characteristics of Byzantine-period
lime mortars and plasters from different construction pe-
riods of the Anaia Church (Kadýkalesi) in Western
Anatolia (Turkey) in order to investigate the continuity
of the lime mortar technology and raw material use
throughout the centuries. The results will also contribute
to future conservation studies at the site.
The Anaia Church surrounded by the Kadýkalesi for-
tification walls was built on a prehistoric mound in the
5–6
th
centuries.
5,6
It had served as the bishopric of the
Ephesus Metropolis between the 5–13
th
centuries and be-
came the archbishopric in the 13
th
century.
5,7–9
The Anaia
Materiali in tehnologije / Materials and technology 56 (2022) 5, 461–470 461
UDK 666.971.3:693.6:94(560Anaia Kadikalesi) ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 56(5)461(2022)
*Corresponding author's e-mail:
tugceisik@iyte.edu.tr (Tuðçe Iþýk)

Church is of great importance as it exhibits an architec-
tural and symbolic value comparable to the monumental
structures from Constantinople (Ýstanbul), the Byzantine
capital city.
Three construction periods were determined in the
Anaia Church, taking into consideration the spatial orga-
nization, construction techniques, materials and architec-
tural elements such as ambo and synthronon (Figure 1).
6
The first construction period (Early Byzantine in the
5–6
th
centuries) with three naves, a naos, a narthex and
atrium spaces covered with a wooden roof was consid-
ered.
6
There had been three major earthquakes that severely
damaged the Church in 1039, 1040 and 1056 with the in-
tensities of grades VII and VIII according to the
MSK-64 intensity scale.
10
Following these earthquakes,
the damaged north and south facade walls of the naos
and narthex were rebuilt and strengthened with but-
tresses during the second construction period between
the 11–13
th
centuries (the Middle Byzantine period).
6,11
The Middle Byzantine period additions can be distin-
guished due to the differences in the wall bonding tech-
niques, recessed brick technique and the separations be-
tween the wall joints. The third construction period was
determined to be between the 13
th
and 14
th
centuries (the
Late Byzantine period) due to sharp wall separations and
different bonding techniques. In this period, the Church
was extended with the baptistery, cisterns and outer
narthex. Besides, the naos walls and buttresses, and inner
narthex buttresses were built to strengthen the Church
structure.
6,11
2 EXPERIMENTAL PART
In this study, 9 lime mortar (M) and 5 lime plaster (P)
samples were taken from the baptistery (Ba), cistern I
(C), cistern II (C), naos (N), outer narthex (O) and sub-
structure (S) of the Anaia Church which were dated to
different construction periods (Figure 1). The sampling
was carried out on the undeteriorated parts of the
Church.
Experimental studies were carried out to determine
the basic physical properties, raw material compositions
and hydraulic properties of the lime mortars and plasters;
mineralogical and chemical compositions, microstruc-
tural properties of binders, aggregates and limes; as well
as pozzolanic activities of the aggregates via the RILEM
test methods, XRD, SEM-EDS and TGA.
The basic physical properties of the lime mortars and
plasters were determined with RILEM tests that de-
scribed their bulk density and porosity values.
12
The raw
material compositions defined by the lime/aggregate ra-
tio and the particle size distributions of the aggregates
were determined by dissolving carbonated lime (CaCO
3
)
with a diluted hydraulic acid (5 %), washing, drying and
sieving the aggregates. The mineralogical compositions
of the finely ground aggregates, binders and limes of less
than 53 μm were detected using X-ray diffraction
(XRD). XRD analyses were done using a Philips X-Pert
Pro X-ray diffractometer with CuK
 radiation, operating
at 40 kV and 40 mA in a range of 5–60° with a scan
speed of 0.08 °/s. The chemical compositions of the ag-
gregates, binders and limes were determined via a scan-
ning electron microscope (SEM) coupled with an X-ray
energy dispersive system (EDS). SEM-EDS analyses
were carried out with a Philips XL 30S FEG on pellets
obtained from powder samples pressed with a pressure
of 980.665 MPa (10 t/cm
2
). Semi-quantitative results for
the chemical compositions were obtained by averaging
the data derived from three distinct areas of the samples
following the k-ratio protocol. The data were collected
without using a standard sample. The results were also
T. IÞÝK, E. U. SAÐÝN: CHARACTERISTICS OF BYZANTINE-PERIOD LIME MORTARS ...
462 Materiali in tehnologije / Materials and technology 56 (2022) 5, 461–470
Figure 1: Plan showing the construction periods of the Anaia Church and sample locations

used to calculate the hydraulic HI (Equation (1)) and ce-
mentation CI (Equation (2)) indices to identify the
hydraulicity of the lime lumps.
13,14
HI =
(%Al O %Fe O %SiO
(%CaO %MgO)
23 23 2
++
+
)
(1)
CI =
(2.8 %SiO 1.1 %Al O 0.7 %Fe O
(%CaO 1.4 %MgO)
22 32 3
⋅+ ⋅+⋅
+⋅
)
(2)
The pozzolanic activities of the aggregates were de-
termined by measuring the differences in electrical con-
ductivity taken before and after the addition of fine ag-
gregates (< 53 μm) into the saturated calcium hydroxide
solution (Ca(OH)
2
).
15
Electrical-conductivity differences
higher than 40 mS/m indicate that aggregates are
pozzolan, while a value of 120 mS/m indicates that ag-
gregates have good pozzolanicity.
15
Pozzolanic activities
were also evaluated with the chemical compositions us-
ing the ASTM C618-03 standard. According to this stan-
dard, the SiO
2
+A l
2
O
3
+F e
2
O
3
content of pozzolanic
materials should be above 70 %.
19
The hydraulic properties of mortars and plasters were
determined by measuring the weight losses due to the
loss of chemically bound water (H
2
O) of hydraulic prod-
ucts between 200–600 °C (H
2
O) and the release of car-
bon dioxide (CO
2
) during the decomposition of CaCO
3
between 600–900 °C (CO
2
) via TGA.
16
Accordingly, a
CO
2
/H
2
O ratio lower than 10 indicates the hydraulic
character of mortar or plaster.
Microstructural properties and morphologies of the
mortars and plasters were specified with a scanning elec-
tron microscope (SEM) coupled with an X-ray energy
dispersive system (EDS) (Philips XL 30S FEG). Analy-
ses were carried out on polished and broken surfaces of
lime mortars using the secondary electron (SE) and
backscattered electron (BSE) modes at different magnifi-
cations in order to examine the general microstructure of
the mortars, properties of pozzolan-binder interfaces and
microstructural characteristics of pozzolans.
3 RESULTS AND DISCUSSION
3.1 Basic physical properties and raw-material compo-
sitions of lime mortars and plasters
Macroscopic investigations revealed that all the lime
mortars consisted of natural aggregates and had a greyish
color, whereas the lime plasters were composed of brick
aggregates and had a pinkish color. All the samples were
sound with a stiff and compact appearance.
The lime mortars with natural aggregates had higher
density and lower porosity of 1.54–1.73 g/cm
3
and
31–37 %, respectively (Table 1). However, the lime plas-
ters with brick aggregates were less dense and more po-
rous with values of 1.30–1.42 g/cm
3
and 43–47 %, re-
spectively (Table 1). These differences may be explained
with the porous structure of brick aggregates.
17,18
It is
known that the porous structure of brick aggregates en-
hances the resistance of plasters to deterioration, making
them ideal materials, especially for water-related struc-
tures.
17
Lime/aggregate ratios by weight were found to be be-
tween 1/3–4/3 for the plasters, 1/2–1/1 for the Early
Byzantine lime mortars, and between 4/3–5/3 for the
Middle and Late Byzantine buttress mortars (Table 1).
These differences might be due to either the use of lime
as the binder or the existence of calcareous based parti-
cles in aggregates. Previous studies showed that a higher
amount of the binder provides better bonding between
the structural components and a greater mechanical
strength in lime mortars.
20–24
The buttresses were thought
to be added to the Anaia Church after the earthquakes
that took place in the Middle and Late Byzantine peri-
ods.
6
The mortars used in the construction of the but-
tresses may have been produced with a higher lime/ag-
gregate ratio to provide a higher strength to resist future
earthquakes.
The particle-size distribution of aggregates is another
important feature in defining the raw-material composi-
tions of lime mortars and plasters since it is known that
T. IÞÝK, E. U. SAÐÝN: CHARACTERISTICS OF BYZANTINE-PERIOD LIME MORTARS ...
Materiali in tehnologije / Materials and technology 56 (2022) 5, 461–470 463
Table 1: Physical properties and raw-material compositions of the lime mortars and plasters
Period Sample Location Function
Aggregate
type
Density
(g/cm
3
)
Porosity
(%)
Lime/aggregate
(by weight)
Early Byzantine
BaM1 Baptistery wall Mortar Natural 1.72 31 0.78
BaM2 Baptistery wall Mortar Natural 1.63 36 1.10
NM2 Naos wall Mortar Natural 1.66 34 0.86
SM1 Substructure arch Mortar Natural 1.62 37 1.14
SM2 Substructure buttress Mortar Natural 1.72 32 0.51
SM3 Substructure vault Mortar Natural 1.58 37 1.14
Middle Byzantine
NM1 Naos buttress Mortar Natural 1.67 35 1.44
NM3 Naos buttress Mortar Natural 1.73 31 1.66
OP1 Outer narthex wall Plaster Brick 1.31 43 1.15
Late Byzantine
OP2 Outer narthex wall Plaster Brick 1.42 44 1.04
CP1 Cistern I wall Plaster Brick 1.34 48 0.72
CP2 Cistern I wall Plaster Brick 1.36 44 1.38
CP3 Cistern II wall Plaster Brick 1.30 46 0.37
NM4 Naos buttress Mortar Natural 1.54 34 1.36

the particle sizes of aggregates effect the physical prop-
erties, durability and mechanical strength of mortars.
The natural aggregates of Kadýkalesi had higher amounts
of aggregates with sizes of > 1180 μm (4–37 %), and
also aggregates of 1180–500 μm (8–21 %), 500–250 μm
(5–32 %), 250–125 μm (3–14 %), 125–53 μm (2–7 %)
by weight, and lower amounts of aggregates of < 53 μm
(0.5–2 %) by weight. Within the distribution of brick ag-
gregates, the weight percentages of the aggregates of the
sizes > 1180 μm (10–19 %), 500–250 μm (9–17 %),
1180–500 μm (4–12 %) were lower; and the weight per-
centages of the aggregates with the sizes of 250–125 μm
(5–12 %), 125–53 μm (6–12 %) and < 53 μm (0.5–5 %)
were higher when compared with the distribution of nat-
ural aggregates. Consequently, the aggregates had a wide
range of particle sizes, enhancing the mechanical
strength of the mortars.
3.2 Characteristics of the lime used in the production
of mortars and plasters
White nodules of a few millimeters in the lime mor-
tars and plasters were considered as "lime lumps", which
might have occurred during the mixing of lime and ag-
gregate.
25
Since lime lumps were considered to represent
lime, their chemical and mineralogical compositions
were accepted to be the same as those of the raw mate-
rial.
25–27
In the XRD patterns of the lime lumps from all
the periods, only sharp calcite peaks were identified
(Figure 2c). The SEM-EDS analysis indicated that lime
lumps mainly consisted of large amounts of CaO
(97–98 %) and smaller amounts of SiO
2
(0.6–1.2 %),
MgO (0.6–0.9 %), Na
2
O (0–0.1 %), Al
2
O
3
(0.4–0.6 %),
K
2
O (0–0.2 %) and were without TiO
2
or Fe
2
O
3
.
Hydraulic and cementation indices were calculated
using the chemical compositions according to the
Boynton formula. Hydraulic-index (HI) values lower
than 0.1, and cementation-index (CI) values lower than
0.3 demonstrate a non-hydraulic character of lime. HIs
and CIs of the lime lumps of the Anaia Church mortars
were found to be between 0.01–0.02 and 0.03–0.04, re-
spectively.
13,14
The mineralogical and chemical composi-
tions and HI and CI values showed that the lime used in
the production is non-hydraulic and fat.
13,14
Micritic cal-
cite crystals smaller than 5 μm were observed in SEM
images (Figures 2a and 2b). This may indicate that the
lime had been used after a long aging.
28
3.3 Characteristics of the natural and brick aggregates
used in mortars and plasters
The pozzolanic activities of the aggregates were de-
termined by following the electrical conductivity differ-
ences in the saturated calcium hydroxide solution before
and after the addition of fine samples (< 53 μm).
15
Elec-
trical conductivity differences were found to be between
150–800 mS/m for the natural aggregates, and between
570–709 mS/m for the brick aggregates (Table 2). The
T. IÞÝK, E. U. SAÐÝN: CHARACTERISTICS OF BYZANTINE-PERIOD LIME MORTARS ...
464 Materiali in tehnologije / Materials and technology 56 (2022) 5, 461–470
Figure 2: Typical lime lump (SM2 sample) consisting of micritic calcite crystals: a) SEM 5000×, b) SEM 20000×, c) XRD pattern (C: calcite
86-2334)

T. IÞÝK, E. U. SAÐÝN: CHARACTERISTICS OF BYZANTINE-PERIOD LIME MORTARS ...
Materiali in tehnologije / Materials and technology 56 (2022) 5, 461–470 465
Figure 3: TAS diagram showing geochemical origins of the fine natural aggregates
Table 2: Chemical and mineralogical compositions and electrical-conductivity differences (E.C.D.) of the aggregates
Period
Sample
Major oxide compositions (%) Mineralogical compositions
Pozzolanic
activity
Na2
O MgO Al
2
O
3
SiO
2
K
2
O CaO TiO
2
Fe
2
O
3 QAMCOH
E.C.D.
(mS/m)
SiO2
+
Al
2
O
3
+
Fe
2
O
3
(%)
Early Byzantine
BaM1
0.60 ±
0.08
2.29 ±
0.29
11.26
± 0.72
75.77
± 1.55
2.25 ±
0.10
1.55
± 0.14
0.84
± 0.05
5.44
± 0.43
+++ ++ + + - - 713 92.47
BaM2
1.86 ±
0.42
2.23
± 0.44
12.71
± 0.72
73.47
± 1.70
2.77
± 0.27
1.81
± 0.30
0.76
± 0.84
4.39
± 0.75
+++ + ++ + ++ - 797 90.57
NM2
0.62 ±
0.11
4.21
± 0.03
16.07
± 0.39
66.90
± 0.24
3.20
± 0.04
0.74
± 0.08
1.04
± 0.13
7.23
± 0.23
+++ ++ ++ ++ - - 553 90.20
SM1
0.66 ±
0.12
4.39
± 0.42
15.47
± 0.54
66.80
± 1.08
3.19
± 0.06
1.06
± 0.05
0.94
± 0.12
7.48
± 0.45
+++ ++ ++ - - - 151 89.75
SM2
2.34 ±
0.41
1.80
± 0.15
16.87
± 1.32
69.53
± 1.20
2.83
± 0.50
0.63
± 0.13
0.71
± 0.56
5.29
± 1.02
+++ ++ ++ - - - 164 91.69
SM3
0.63 ±
0.16
3.02
± 0.33
12.54
± 0.15
74.38
± 0.65
2.79
± 0.19
0.83
± 0.07
0.64
± 0.22
5.18
± 0.11
+++ ++ ++ ++ ++ - 803 92.10
Middle
Byzantine
NM1
0.78 ±
0.11
3.27
± 0.19
14.11
± 0.35
69.74
± 0.67
3.40
± 0.07
0.99
± 0.07
1.22
± 0.21
6.51
± 0.65
+++ ++ ++ ++ - - 433 90.36
NM3
0.80
± 0.16
3.02
± 0.40
13.56
± 0.83
71.53
± 1.18
3.11
± 0.10
1.17
± 0.13
1.00
± 0.18
5.82
± 0.17
+++ ++ ++ + - - 790 90.91
OP1
0.99
± 0.12
1.35
± 0.14
13.60
± 0.30
73.41
± 0.66
2.89
± 0.23
0.83
± 0.09
0.80
± 0.24
6.13
± 0.23
+++ ++ ++ - - + 757 93.14
Late Byzantine
OP2
0.99
± 0.21
2.08
± 0.14
16.76
± 0.13
67.20
± 0.49
3.76
± 0.17
1.23
± 0.06
0.73
± 0.41
7.26
± 0.32
+++ ++ ++ - - + 568 91.22
CP1
0.52
± 0.12
2.07
± 0.10
10.41
± 0.27
78.15
± 0.52
2.13
± 0.05
2.09
± 0.23
0.58
± 0.10
4.04
± 0.54
+++ ++ ++ - - + 769 92.60
CP2
0.49
± 0.08
2.98
± 0.18
14.40
± 0.13
66.12
± 0.19
3.04
± 0.20
4.10
± 0.14
1.03
± 0.06
7.83
± 0.31
+++ ++ ++ - - - 785 88.35
CP3
0.47
± 0.13
3.07
± 0.12
12.06
± 0.70
71.90
± 0.53
2.60
± 0.10
2.96
± 0.16
0.93
± 0.10
6.01
± 0.38
+++ ++ ++ - - + 785 89.97
NM4
0.37
± 0.07
2.24
± 0.14
15.27
± 0.17
71.14
± 0.39
3.27
± 0.14
0.85
± 0.06
1.46
± 0.30
5.40
± 0.55
+++ ++ ++ + - - 308 91.81
The number of pluses represents the abundance of mineral peaks. (Q: quartz, A: albite, M: muscovite, Cl: clinochlore,
O: orthoclase, H: hematite)

results greater than 120 mS/m revealed that both the nat-
ural and brick aggregates used in the mortars and plasters
from all periods exhibited highly reactive pozzolanic
properties.
15
SEM-EDS analyses revealed that the natural aggre-
gates in lime mortars were mainly composed of large
amounts of SiO
2
(66.8–75.8 %), moderate amounts of
Al
2
O
3
(11.3–16.9 %) and smaller amounts of Fe
2
O
3
(4.4–7.4 %), MgO (1.8–4.4 %), K
2
O (2.2–3.4 %), Na
2
O
(0.4–2.3 %), CaO (0.6–1.8 %) and TiO
2
(0.6–1.5 %) (Ta-
ble 2). The SiO
2
+Al
2
O
3
+Fe
2
O
3
content of the natural
aggregates was in a range of 89.75–92.47, indicating
their pozzolanic property.
19
Possible geochemical classes of fine natural aggre-
gates (< 53 μm) were evaluated using the total alkali-sil-
ica (TAS) diagram according to the major oxide compo-
sitions.
29
According to the TAS diagram, all fine natural
aggregates from the Middle and Late Byzantine period
and some from the Early Byzantine periods (BaM1,
NM2, SM1, SM2) were classified as the dacite group,
while two of the Early Byzantine samples (BaM2, SM3)
were in the rhyolite group (Figure 3). The fine natural
aggregates from the dacite group were composed of
quartz, albite, muscovite and clinochlore, while the fine
natural aggregates from the rhyolite group also contained
orthoclase (Figures 4b and 4d), (Table 2).
The use of volcanic aggregates was a common prac-
tice in the ancient Hellenistic, Roman and Byzantine set-
tlements in western Anatolia that has rich igneous rock
resources.
18,30–32
Future interdisciplinary studies may pro-
vide more precise information about the provenance of
natural aggregates. Determining the sources will also be
necessary for the mortars to be produced for the conser-
vation works in the future.
The chemical compositions of the brick aggregates
used in the Middle and Late Byzantine lime plasters
mainly consisted of large amounts of SiO
2
(66.1–78.2 %), moderate amounts of Al
2
O
3
(10.4–16.8 %) and smaller amounts of Fe
2
O
3
(4.0–7.8 %), MgO (1.4–3.1 %), K
2
O (2.1–3.8 %), Na
2
O
(0.5–1.0 %), CaO (0.8–4.1 %) and TiO
2
(0.6–1.0 %) ac-
cording to SEM-EDS (Table 2). Smaller amounts of
CaO indicated that all the brick aggregates were pro-
duced from Ca-poor clay sources. The SiO
2
+A l
2
O
3
+
Fe
2
O
3
values were found to be in a range of 88.35–93.14,
indicating a pozzolanic property as determined with the
electrical-conductivity analysis.
19
The XRD analysis demonstrated that the brick aggre-
gates were mostly comprised of quartz, albite and mus-
covite. Different mineral phases of the brick aggregates were
used to predict their firing temperatures. At firing tem-
peratures exceeding 900 °C, amorphous substances begin
to disappear and pozzolanic activities are lost. The pres-
T. IÞÝK, E. U. SAÐÝN: CHARACTERISTICS OF BYZANTINE-PERIOD LIME MORTARS ...
466 Materiali in tehnologije / Materials and technology 56 (2022) 5, 461–470
Figure 4: SEM images and XRD diffraction patterns of the aggregates: a) brick SEM 2500×, b) brick XRD, c) natural SEM 2500×, d) natural
XRD (A: albite 76–1819, Cl: clinochlore 79–1270, H: hematite 87–1166, M: muscovite 84–1302, Q: quartz 85–0798)

ence of mineral phases such as mullite ( 1000 °C), crist-
obalite ( 1200 °C) or wollastonite ( 900–1050 °C) in-
dicated high firing temperatures.
33,34
Pozzolanic
properties and the absence of high-temperature minerals
suggested that the firing temperature did not exceed
 900 °C in all the brick aggregates.
Microstructural properties of the aggregates were de-
termined with SEM-EDS. SEM images showed that
brick aggregates had a more porous structure than natu-
ral aggregates. The microstructure of the brick aggre-
gates also exhibited little vitrification, confirming low
firing temperatures during their production. Both aggre-
gates were comprised of amorphous particles with an ir-
regular morphology, which can be associated with their
pozzolanic properties (Figures 4a and 4c). These amor-
phous substances could not be determined with the XRD
patterns since their non-crystalline structure did not give
any indicative peaks.
3.4 Characteristics of the binders of mortars and plas-
ters
Fine mortar and plaster matrices, finer than 63 μm,
consisted of small grain-sized aggregates and carbonated
lime was defined as the "binder".
25,35
Binders are the
parts that provide high strength and hydraulic character-
istic to mortars. The SEM-EDS analysis revealed that the
binders comprised of natural aggregates had a larger
amount of CaO (67.0–87.2 %) and small amounts of
SiO
2
(6.9–20.3 %) and Al
2
O
3
(2.2–6.4 %), while the
binders with brick aggregates consisted of a larger
amount of CaO (36.4–64.9 %) and moderate amounts of
SiO
2
(19.5–34.5 %) and Al
2
O
3
(9.2–14.1 %) (Table 3).
In the XRD diffraction patterns of the binders with
natural or brick aggregates, calcite originating from lime;
and quartz, albite and muscovite originating from aggre-
gates were determined (Figures 5b and 6b), (Table 3).
Mineralogical compositions of the binders indicated sim-
ilar characteristics throughout different periods. Hydrau-
lic products like calcium silicate hydrates (CSH) and cal-
cium aluminate hydrates (CAH) formed as the result of
the reaction between the pozzolanic aggregates and lime
binder could not be detected due their amorphous struc-
ture.
Hydraulic properties of the binders were determined
by calculating the percentages of weight losses between
200–600 °C and 600–900 °C, detected via TGA. The
weight losses between 200–600 °C were due to structur-
T. IÞÝK, E. U. SAÐÝN: CHARACTERISTICS OF BYZANTINE-PERIOD LIME MORTARS ...
Materiali in tehnologije / Materials and technology 56 (2022) 5, 461–470 467
Table 3: Chemical, mineralogical compositions and hydraulic properties of the binders
Period
Sample
Major oxide compositions (%) Mineralogical compositions
Hydraulic
properties
Na2
O MgO Al
2
O
3
SiO
2
K
2
O CaO TiO
2
Fe
2
O
3 CQAM
CO
2
H
2
O
Early Byzantine
BaM1
0.22
± 0.11
3.65
± 0.36
6.18
± 0.92
19.08
± 1.10
1.30
± 0.15
66.95
± 1.04
0.35
± 0.18
2.28
± 0.23
+++ ++ + + 5.8
BaM2
0.00
± 0.13
1.41
± 0.09
2.69
± 0.87
7.45
± 0.05
0.65
± 0.15
86.47
± 0.44
0.05
± 0.05
1.28
± 0.36
+++ ++ + + 9.1
NM2
0.33
± 0.03
2.47
± 0.36
6.40
± 0.08
20.25
± 1.45
1.16
± 0.18
67.14
± 0.12
0.35
± 1.98
1.90
± 0.10
+++ ++ + + 5.2
SM1
0.46
± 0.21
1.95
± 0.08
2.83
± 0.18
9.45
± 0.34
0.87
± 0.09
83.54
± 1.03
0.00
± 0.00
0.90
± 0.25
+++ ++ + + 8.5
SM2
0.43
± 0.16
2.68
± 0.26
4.94
± 2.18
16.12
± 0.92
1.25
± 0.11
72.50
± 1.39
0.16
± 0.17
1.92
± 0.20
+++ ++ + + 7.2
SM3
0.29
± 0.16
1.91
± 0.17
2.19
± 0.62
6.85
± 0.65
0.37
± 0.18
87.24
± 0.52
0.12
± 0.06
1.02
± 0.29
+++ ++ + + 8.1
Middle
Byzantine
NM1
0.23
± 0.02
1.40
± 0.19
2.99
± 0.19
9.69
± 0.59
0.44
± 0.11
84.53
± 1.25
0.00
± 0.00
0.72
± 0.44
+++ ++ + + 10.0
NM3
0.42
± 0.20
2.01
± 0.26
5.13
± 0.80
18.95
± 1.18
0.85
± 0.15
71.51
± 0.56
0.00
± 0.00
1.13
± 0.28
+++ ++ + + 8.3
OP1
0.42
± 0.07
6.67
± 0.01
14.05
± 0.18
34.51
± 0.28
2.48
± 0.11
36.42
± 0.64
0.58
± 0.16
4.87
± 0.37
+++ ++ + + 3.4
Late Byzantine
OP2
0.61
± 0.14
5.66
± 0.16
11.20
± 1.60
32.34
± 0.91
2.53
± 0.08
42.56
± 1.81
0.40
± 0.11
4.70
± 0.36
+++ ++ + + 2.7
CP1
0.47
± 0.15
3.28
± 0.18
12.82
± 0.03
33.21
± 0.26
2.31
± 0.02
43.30
± 0.35
0.53
± 0.14
4.07
± 0.22
+++ ++ + + 2.7
CP2
0.70
± 0.04
3.30
± 0.10
12.49
± 0.59
29.57
± 0.03
2.32
± 0.03
46.81
± 0.43
0.42
± 0.15
4.37
± 0.12
+++ ++ + + 4.8
CP3
0.50
± 0.06
1.80
± 0.17
9.15
± 0.65
19.49
± 0.74
1.18
± 0.02
64.86
± 0.30
0.43
± 0.13
2.59
± 0.24
+++ ++ + + 3.1
NM4
0.21
± 0.25
1.40
± 0.04
3.62
± 0.12
13.56
± 1.33
0.69
± 0.06
79.22
± 1.20
0.24
± 0.14
1.05
± 0.11
+++ ++ + + 8.4
The number of pluses represents the abundance of mineral peaks. (C: calcite, Q: quartz, A: albite, M: muscovite)

ally bound water (H
2
O) of the hydraulic products (CSH,
CAH), while the losses between 600–900 °C were
caused by a CO
2
release during the decomposition of
carbonate lime. The CO
2
/H
2
O ratio between 1 and 10 in-
dicated the hydraulic character of the mortars.
16
The
CO
2
/H
2
O ratios were found to be between 5.2–10.0 by
weight for the mortars composed of natural aggregates
and between 2.7–4.8 by weight for the lime plasters
composed of brick aggregates (Table 3). These results
revealed that all the binders exhibited hydraulic proper-
ties. The hydraulicity of the binders may be attributed to
the use of highly reactive natural and artificial pozzo-
lanic aggregates in their production.
SEM images of the binders showed strong bonding
between the lime and aggregates, which was an indicator
of a pozzolanic reaction and hydraulic properties. Also,
there was no microcrack formation or irregularity at the
interfaces (Figures 5a and 6a)( Tables 4 and 5).
T. IÞÝK, E. U. SAÐÝN: CHARACTERISTICS OF BYZANTINE-PERIOD LIME MORTARS ...
468 Materiali in tehnologije / Materials and technology 56 (2022) 5, 461–470
Figure 5: Binder (B) with a brick aggregate (B.A.) – sample OP1: a) SEM image 250×, b) XRD pattern (A: albite 76–1819, C: calcite 86–2334,
M: muscovite 84–1302, Q: quartz 85–0798)
Table 4: Major oxide compositions of the binder and brick aggregate in sample OP1
Major oxide compositions (w/%)
Na2
O MgO Al
2
O
3
SiO
2
K
2
O CaO TiO
2
Fe
2
O
3
Binder 0.61 4.32 5.89 17.65 0.88 68.35 0.10 2.20
Brick aggregate 1.13 4.15 17.35 38.94 3.36 29.58 0.48 5.02
Figure 6: Binder (B) with a natural aggregate (N.A.) – sample NM2: a) SEM image 500×, b) XRD pattern (A: albite 76–1819, C: calcite
86–2334, M: muscovite 84–1302, Q: quartz 85–0798)
Table 5: Major oxide compositions of the binder and natural aggregate in sample NM2
Major oxide compositions (w/%)
Na2
O MgO Al
2
O
3
SiO
2
K
2
O CaO TiO
2
Fe
2
O
3
Binder 0.56 1.07 2.84 21.94 1.46 71.50 0.13 0.50
Natural aggregate 0.11 0.22 0.46 95.09 0.27 3.86 0.00 0.00

4 CONCLUSIONS
Byzantine-period lime mortars and plasters used in
different construction phases of the Anaia Church are
stiff and compact materials that survived for centuries,
preserving their original characteristics. All the lime
mortars were produced from natural aggregates of a vol-
canic origin with lime/aggregate ratios between 1/2–5/3,
whereas all the lime plasters consisted of brick aggre-
gates with lime/aggregate ratios between 1/3–4/3. The
only exception was the mortars used for the buttresses
added to the church to strengthen its structure after the
earthquakes in the 11
th
century, with a higher lime/aggre-
gate ratio of 4/3–5/3.
Non-hydraulic and fat lime had been used for the
mortar and plaster production after a long aging period.
Natural aggregates of mortars were probably mainly ob-
tained from the local rhyolite and dacite sources. Brick
aggregates were manufactured from Ca-poor clays at low
firing temperatures between 800–900 °C. All the natural
and brick aggregates used in the mortars and plasters ex-
hibited highly reactive pozzolanic properties. Due to the
use of pozzolanic aggregates in their production, mortars
and plasters had hydraulic characteristics.
It is remarkable that the basic physical properties,
raw-material compositions, chemical and mineralogical
compositions of the lime mortars and plasters used in
different construction periods of the Anaia Church,
which is a Byzantine structure, were similar and have not
changed over the centuries. These similarities reveal that
the knowledge of the use of the local raw-material re-
sources and the mortar production technology was inten-
tionally transferred over the centuries during the
Byzantine period.
The production of new lime mortars and plasters to
be used in the future conservation works in the Anaia
Church should be physically, chemically and mineralogi-
cally compatible with the original mortar and plaster
properties determined by this study. For this purpose,
possible local lime, natural pozzolan and clay sources
should be investigated and their suitability for the use in
the mortar and plaster production should be fully exam-
ined. Furthermore, special production of brick aggre-
gates should be carried out and these aggregates should
be fired at low temperatures.
Acknowledgement
The authors would like to thank Prof. Dr. Zeynep
Mercangöz and excavation team for their support in col-
lecting samples from the ancient site of Kadýkalesi
(Anaia). The authors also thank the Integrated Research
Centers at the Ýzmir Institute of Technology for
SEM-EDS, XRD and TGA analyses during the experi-
mental phase of this investigation.
5 REFERENCES
1
N. A. Davey, History of Building Materials, Phoenix House, London
1961
2
J. A. H. Oates, Lime and Limestone: Chemistry and Technology,
Production and Uses, Wiley-VCH, Weinheim 1998
3
H. Böke, S. Akkurt, B. Ipekoðlu, E. Uðurlu, Characteristics of Brick
Used as Aggregate in Historic Brick-Lime Mortars and Plasters,
Cem. Concr. Res., 36 (2006), 1115–1122, doi:10.1016/j.cemconres.
2006.03.011
4
J. P. Adam, Roman Building Materials and Techniques, (1
st
pub. in
1937), Routledge, London and New York 2005
5
Z. Mercangöz, Kuºadasý,K a d ýkalesi (Anaia), 1st ed., Ege
Üniversitesi Yayýnlarý, Ýzmir 2012
6
M. B. Kanmaz, B. Ýpekoðlu, Restorations due to Earthquakes in the
Ancient Cities: Anaia Byzantine Church, Kargir Yapýlarda Koruma
ve Onarým Semineri VIII, 2016, 189–205
7
C. Foss, Ephesus after Antiquity: A Late Antique, Byzantine and
Turkish City, Cambridge University Press, New York 1979
8
P. Culerrier, Les évêchés suffragants d’Éphèse aux 5e-13e siècles,
Rev. Etud. Byz., 45 (1987) 1, 139–164, doi:10.3406/rebyz.1987.2207
9
Z. Mercangöz, The Variable Historical Destiny of Anaia as
Emporion and Kommerkion, I. Uluslararasý Sevgi Gönül Bizans
Araºtýrmalarý Sempozyumu, 2010, 279–292
10
H. Soysal, S. Sipahioðl u ,D .K o l ç a k ,Y .A l t ýnok, Historical Earth-
quake Catalog of Turkey and its Surroundings (BC 2100–AC 1900),
Tübitak Yayýnlarý, Ýstanbul 1981
11
M. B. Kanmaz, Evalation of Conservation Problems of Anaia Byzan-
tine Church, Kadýkalesi, Kuºadasý, Izmir Institute of Technology,
2015
12
RILEM:1980 – Tests Defining the Structure, Vol. 13, No. 73
13
E. C. Eckel, Cements, Limes, and Plasters: Their Materials, Manu-
facture, and Properties, 1
st
ed., John Wiley & Sons, New York 1905
14
R. S. Boynton, Chemistry and Technology of Lime and Limestone,
John Wiley & Sons, 1966
15
M. P. Luxan, F. Madruga, J. Saavedra, Rapid Evaluation of
Pozzolanic Activity of Natural Products by Conductivity Measure-
ment, Cem. Concr. Res., 19 (1989) 1, 63–68, doi:10.1016/0008-
8846(89)90066-5
16
A. Bakolas, G. Biscontin, A. Moropoulou, E. Zendri, Characteriza-
tion of Structural Byzantine Mortars by Thermogravimetric Analysis,
Thermochim. Acta, 321 (1998), 151–160, doi:10.1016/s0040-
6031(98)00454-7
17
E. Uðurlu, H. Böke, The Use of Brick–Lime Plasters and Their Rele-
vance to Climatic Conditions of Historic Bath Buildings, Constr.
Build. Mater., 23 (2009) 6, 2442–2450, doi:10.1016/j.conbuildmat.
2008.10.005
18
Ö. Aslan Özkaya, H. Böke, Properties of Roman Bricks and Mortars
Used in Serapis Temple in the City of Pergamon, Mater. Charact, 60
(2009) 9, 995–1000, doi:10.1016/j.matchar.2009.04.003
19
ASTM C618-03, Standard Specification for Coal Fly Ash and Raw
or Calcined Natural Pozzolan for Use in Concrete (2003)
20
J. Lanas, J. I. Alvarez, Masonry Repair Lime-Based Mortars: Factors
Affecting the Mechanical Behavior, Cem. Concr. Res, 33 (2003) 11,
1867–1876, doi:10.1016/S0008-8846(03)00210-2
21
A. El-Turki, R. J. Ball, M. A. Carter, M. A. Wilson, C. Ince, G. C.
Allen, Effect of Dewatering on the Strength of Lime and Cement
Mortars, J. Am. Ceram. Soc., 93 (2010) 7, 2074–2081, doi:10.1111/
j.1551-2916.2010.03667.x
22
E. Aggelakopoulou, A. Bakolas, A. Moropoulou, Properties of
Lime-Metakolin Mortars for the Restoration of Historic Masonries,
Appl. Clay Sci., 53 (2011)1, 15–19, doi:10.1016/j.clay.2011.04.005
23
A. Gameiro, A. Santos Silva, P. Faria, J. Grilo, T. Branco, R. Veiga,
A. Velosa, Physical and Chemical Assessment of Lime-Metakaolin
Mortars: Influence of Binder:Aggregate Ratio, Cem. Concr. Compos,
45 (2014), 264–271, doi:10.1016/j.cemconcomp.2013.06.010
T. IÞÝK, E. U. SAÐÝN: CHARACTERISTICS OF BYZANTINE-PERIOD LIME MORTARS ...
Materiali in tehnologije / Materials and technology 56 (2022) 5, 461–470 469

24
A. M. S. Loureiro, S. P. A. da Paz, M. R. do Veiga, R. S. Angélica,
Investigation of Historical Mortars from Belém Do Pará, Northern
Brazil, Constr. Build. Mater., 233 (2020), doi:10.1016/j.conbuildmat.
2019.117284
25
A. Bakolas, G. Biscontin, A. Moropoulou, E. Zendri, Characteriza-
tion of the Lumps in the Mortars of Historic Masonry, Thermochim.
Acta, 269 (1995), 809–816, doi:10.1016/0040-6031(95)02573-1
26
S. Bruni, F. Cariati, P. Fermo, P Cairati, G. Alessandrini, L. Toniolo,
White Lumps in Fifth- to Seventeenth-Century AD Mortars from
Northern Italy, Archaeometry, 39 (1997) 1, 1–7, doi:10.1111/j.1475-
4754.1997.tb00786.x
27
L. Barba, J. Blancas, L. R. Manzanilla, A. Ortiz, D. Barca, G. M.
Crisci, D. Miriello, A. Pecci, Provenance of the Limestone Used in
Teotihuacan (Mexico): A Methodological Approach, Archaeometry,
51 (2009) 4, 525–545, doi:10.1111/j.1475-4754.2008.00430.x
28
I. C. Zamba, M. G. Stamatakis, F. A. Cooper, P. G. Themelis, C. G.
Zambas, Characterization of Mortars Used for the Construction of
Saithidai Heroon Podium (1st Century AD) in Ancient Messene,
Peloponnesus, Greece, Mater. Charact, 58 (2007), 1229–1239,
doi:10.1016/j.matchar.2007.07.004
29
R. W. Le Maitre, A. Streckeisen, B. Zanettin, M. J. Le Bas, B. Bonin,
P. Bateman, G. Bellieni, A. Dudek, S. Efremova, J. Keller, J.
Lameyre, P. A Sabine, R. Schmid, H. Sørensen, A. R. Woolley, Igne-
ous Rocks. A Classification and Glossary of Terms, 2nd ed., Cam-
bridge University Press, New York 2002
30
D. Miriello, A. Bloise, G. M. Crisci, C. Apollaro, A. La Marca,
Characterisation of Archaeological Mortars and Plasters from Kyme
(Turkey), J. Archaeol. Sci., 38 (2011) 4, 794–804, doi:10.1016/j.jas.
2010.11.002
31
C. Helvaci, Geological Features of Neogene Basins Hosting Borate
Deposits: An Overview of Deposits and Future Forecast, Turkey,
Bull. Miner. Res. Explor., 151 (2015), 169–215, doi:10.19111/bmre.
05207
32
E. Uðurlu Saðýn, H. E. Duran, H. Böke, Lime Mortar Technology in
Ancient Eastern Roman Provinces, J. Archaeol. Sci. Reports, 39
(2021), 1–14, doi:10.1016/j.jasrep.2021.103132
33
P. Cardiano, S. Ioppolo, C. De Stefano, A. Pettignano, S. Sergi, P.
Piraino, Study and Characterization of the Ancient Bricks of Monas-
tery of "San Filippo Di Fragalà" in Frazzanò (Sicily), Anal. Chim.
Acta, 519 (2004), 103–111, doi:10.1016/j.aca.2004.05.042
34
S. Lee, Y. J. Kim, H. S. Moon, Phase Transformation Sequence from
Kaolinite to Mullite Investigated by an Energy-Filtering Transmis-
sion Electron Microscope, Journal of the American Ceramic Society,
82 (1999) 10, 2841–2848, doi:10.1111/j.1151-2916.1999.tb02165.x
35
B. Middendorf, J. J. Hughes, K. Callebaut, G. Baronio, I. Papayianni,
Investigative Methods for the Characterisation of Historic Mortars –
Part 2, Chemical Characterisation, Mater. Struct., 38 (2005) 282,
771–780, doi:10.1617/14282
T. IÞÝK, E. U. SAÐÝN: CHARACTERISTICS OF BYZANTINE-PERIOD LIME MORTARS ...
470 Materiali in tehnologije / Materials and technology 56 (2022) 5, 461–470