Physical–Mechanical and Mineralogical Properties of Fired Bricks of the Archaeological Site of Harran, Turkey

In this study, the physical–mechanical and mineralogical properties of bricks used in historical structures of the site of Harran, Turkey have been investigated. Harran was destroyed by the Mongol army, during the Turkish reconquest campaign around the 1260s. The remains of the buildings made of bricks and basalt/limestone were recently uncovered almost in their entirety. Several brick samples have been taken from the burial mound, the university, the city walls, the castle, and the great mosque. From the visual analyses, it was noted that the bricks have unique colors such as pottery, desert beige, and canyon. Physical analyses show that the absorption rates of the bricks are between 17.30–38.12%, their densities between 1.33–1.70 g/cm3, and porosities between 8.88–25.31%. For the mechanical analyses, the bending strengths have been found to be between 0.82–1.86 MPa and compressive strengths between 6.69–7.95 MPa. The thin-section images show that the bricks contain calcite, muscovite, and plagioclase quartz and pyroxene minerals.


Introduction
Conservation, preservation, and restoration of cultural heritage assets and buildings is an emerging topic nowadays. Protection of the heritage created over the thousand-year history of human civilization has become urgent also, considering the increasing numbers of natural and man-made hazards [1][2][3].
It is a challenging topic and requires a multidisciplinary approach, involving experts of different fields. The implementation of guidelines and procedures was set by the early 20th century, such as Carta del Restauro (Athens Charter for the Restoration of Historic Monuments) [4] in 1931, the Venice Charter [5] in 1964, and in ICOMOS (International Council on Monuments and Sites) concerning research, documentation, and technical support regarding cultural assets [6,7]. The final document of the 1974 UNESCO conference [8] and the European Council guidelines [9] regarding the protection of Architectural Heritage represented a keystone in this area. Recently, new methods and procedures have been developed to study historical constructions by means of visual inspection, in-situ testing, (in that period the city was known as Eλληνóπoλις or Kάρραι) and, finally, in 750 AD, the Arabs invading the region ended the Byzantine dominance. After the settlement of the Turkish states of Fatimids, Zengids, Ayyubids, and Seljuks [41], the city was conquered by the Mongols around the 1260s. In this long period of time, the city flourished and was a major urban agglomerate in the region. The city and its monuments are cited in a large number of old documents and historic records. When the Mongols were expelled in 1270 by the Turkish army, the city was partially destroyed: the mosque, the city walls, and the castle were highly damaged by the Mongols.
The most important monuments of the city are: Burial Mound: This monument which has partially survived until today is located in the center of the historic site of Harran, has a height of 22 m, and is spread over a fairly large area. This monument has been continually used from the 3rd century BC until the 13th century AD. The masonry ruins it contains belong to various periods. It is worth noting that several samples of cuneiform scripts were found here, dating its construction to the Sumerian period. According to a hypothesis difficult to prove at the moment, the burial mound should be identified with a moon temple. A moon cult ruled the Plains of Harran for several centuries, before the diffusion of Christianity and Islam. More evidence about this monument could be found in the near future, also considering that the burial mound is actually an active excavation site [42].
Castle: The exact date of construction of this masonry structure is not known. Recently, an ancient Greek inscription was found near the south-east gate: this could help with dating (the Byzantine period of the fourth-seventh century). However, it is likely that, in the ninth century, the Castle, after the conquest of the Arabs (seventh century), was enlarged and restored. According to professor Önal from Harran University [42], this was initially a palace, and it was turned into a castle-like structure with polygonal towers during the turbulent periods of invasions of Mongols and the Crusades (11-13th centuries). In fact, the city of Harran was near the border with the Christian state of the County of Edessa (todayŞanlıurfa in Turkey). According to the historic records, the Castle was heavily damaged and almost destroyed during the Mongolian invasions (13th century).
The site is in the process of being added to UNESCO's World Heritage Temporary List [43]. Recently, excavations have been carried out at the castle and nearby buildings with the support of the Turkish Ministry of Culture and Tourism. The excavations carried out on the second floor of the Palace revealed a three-domed bath: the cold, warm, and furnace sections of the bath were completely uncovered. Part of the Castle was recently re-constructed using similar stonework ( Figure 1). The archeological site of Harran is not yet fully open to public: according to professor Önal [42] the site was supposed to be accessible at the end of 2019, but the Covid-19 pandemic has likely delayed this.
Parthians defeated the Romans under the command of Crassus, who was killed. The city passed under the control of the Eastern Roman Empire (in that period the city was known as Ἑλληνόπολις or Κάρραι) and, finally, in 750 AD, the Arabs invading the region ended the Byzantine dominance. After the settlement of the Turkish states of Fatimids, Zengids, Ayyubids, and Seljuks [41], the city was conquered by the Mongols around the 1260s. In this long period of time, the city flourished and was a major urban agglomerate in the region. The city and its monuments are cited in a large number of old documents and historic records. When the Mongols were expelled in 1270 by the Turkish army, the city was partially destroyed: the mosque, the city walls, and the castle were highly damaged by the Mongols.
The most important monuments of the city are: Burial Mound: This monument which has partially survived until today is located in the center of the historic site of Harran, has a height of 22 m, and is spread over a fairly large area. This monument has been continually used from the 3rd century BC until the 13th century AD. The masonry ruins it contains belong to various periods. It is worth noting that several samples of cuneiform scripts were found here, dating its construction to the Sumerian period. According to a hypothesis difficult to prove at the moment, the burial mound should be identified with a moon temple. A moon cult ruled the Plains of Harran for several centuries, before the diffusion of Christianity and Islam. More evidence about this monument could be found in the near future, also considering that the burial mound is actually an active excavation site [42].
Castle: The exact date of construction of this masonry structure is not known. Recently, an ancient Greek inscription was found near the south-east gate: this could help with dating (the Byzantine period of the fourth-seventh century). However, it is likely that, in the ninth century, the Castle, after the conquest of the Arabs (seventh century), was enlarged and restored. According to professor Ö nal from Harran University [42], this was initially a palace, and it was turned into a castle-like structure with polygonal towers during the turbulent periods of invasions of Mongols and the Crusades (11-13th centuries). In fact, the city of Harran was near the border with the Christian state of the County of Edessa (today Şanlıurfa in Turkey). According to the historic records, the Castle was heavily damaged and almost destroyed during the Mongolian invasions (13th century).
The site is in the process of being added to UNESCO's World Heritage Temporary List [43]. Recently, excavations have been carried out at the castle and nearby buildings with the support of the Turkish Ministry of Culture and Tourism. The excavations carried out on the second floor of the Palace revealed a three-domed bath: the cold, warm, and furnace sections of the bath were completely uncovered. Part of the Castle was recently re-constructed using similar stonework ( Figure 1). The archeological site of Harran is not yet fully open to public: according to professor Ö nal [42] the site was supposed to be accessible at the end of 2019, but the Covid-19 pandemic has likely delayed this.
According to the detailed survey by Rice [44], carried out around 1950, the Castle is a threestory building, with tens of rooms. The masonry material is mainly cut stones (basalt and limestone blocks of varying sizes, some of them over a meter long). According to the detailed survey by Rice [44], carried out around 1950, the Castle is a three-story building, with tens of rooms. The masonry material is mainly cut stones (basalt and limestone blocks of varying sizes, some of them over a meter long).  [45], carried out excavation works for restoration purposes on the western part of the city wall, uncovering the city walls, several brickwork towers, and bastions. In excavations in the northern part of the Castle, a gallery was discovered on the west side. It is likely that the construction of the walls dates back to the Roman-Byzantine period: near the south-east gate, an inscription in ancient Greek was found set in a wall. In 2016, excavations were carried out on the city wall (west of the southern, "Raqqa" Gate), revealing part of the wall and leading to the discovery of a broken statue of a woman with a Syriac inscription and a male relief, both used as spolia in the wall. In the 2014-2016 excavations carried out in the west side of the Castle, a crenellated corridor belonging to a second defense system adjacent to the wall of the Castle (between the polygonal and rectangular towers) was uncovered.
Great Mosque: The great mosque is considered the oldest mosque built in Anatolia as part of the Islamic architecture. This mosque is known as the Paradise Mosque. It is a stone and brickwork masonry construction, built by the last Ummayad caliph Mervan II in 744-750. The plan of the building has the dimensions of 104 × 107 m. The mosque was recently unearthed during the excavations led by Dr Nurettin Yardimer [45]. These excavations are currently being carried out also outside the northern and western gates. Only small portions of the mosque remain standing today: the 33.3 m tall minaret, the fountain, the mihrab, and the eastern wall. The masonry members have gone through several restoration processes. The mosque, which is located at the eastern foothill of the burial mound, is also known by the names Cami el-Firdevs and Cuma Camisi (Friday Mosque).
A historic copper map of the region is given in Figure 2. It can be noted that the site of Harran, 44 km southeast ofŞanlıurfa, is located in a strategic region near the historic and modern border with Mesopotamia, Syria, and Persia, under the Southeastern Anatolia mountain range.   [45], carried out excavation works for restoration purposes on the western part of the city wall, uncovering the city walls, several brickwork towers, and bastions. In excavations in the northern part of the Castle, a gallery was discovered on the west side. It is likely that the construction of the walls dates back to the Roman-Byzantine period: near the south-east gate, an inscription in ancient Greek was found set in a wall. In 2016, excavations were carried out on the city wall (west of the southern, "Raqqa" Gate), revealing part of the wall and leading to the discovery of a broken statue of a woman with a Syriac inscription and a male relief, both used as spolia in the wall. In the 2014-2016 excavations carried out in the west side of the Castle, a crenellated corridor belonging to a second defense system adjacent to the wall of the Castle (between the polygonal and rectangular towers) was uncovered.
Great Mosque: The great mosque is considered the oldest mosque built in Anatolia as part of the Islamic architecture. This mosque is known as the Paradise Mosque. It is a stone and brickwork masonry construction, built by the last Ummayad caliph Mervan II in 744-750. The plan of the building has the dimensions of 104 × 107 m. The mosque was recently unearthed during the excavations led by Dr Nurettin Yardimer [45]. These excavations are currently being carried out also outside the northern and western gates. Only small portions of the mosque remain standing today: the 33.3 m tall minaret, the fountain, the mihrab, and the eastern wall. The masonry members have gone through several restoration processes. The mosque, which is located at the eastern foothill of the burial mound, is also known by the names Cami el-Firdevs and Cuma Camisi (Friday Mosque).
A historic copper map of the region is given in Figure 2. It can be noted that the site of Harran, 44 kilometers southeast of Şanlıurfa, is located in a strategic region near the historic and modern border with Mesopotamia, Syria, and Persia, under the Southeastern Anatolia mountain range.

Materials
Five brick samples were taken from different historical ruins of the city: the burial mound, the university, the city walls, the castle and the great mosque, from undisturbed parts of these buildings. While extracting the brick samples, special attention was paid not to spoil the unique structures of the bricks and photos of places from where the samples were taken, and photos of the taken brick samples were filed. The five selected buildings are shown in Figure 3. A detail of a wall of the burial mound is given in Figure 4.

Materials
Five brick samples were taken from different historical ruins of the city: the burial mound, the university, the city walls, the castle and the great mosque, from undisturbed parts of these buildings. While extracting the brick samples, special attention was paid not to spoil the unique structures of the bricks and photos of places from where the samples were taken, and photos of the taken brick samples were filed. The five selected buildings are shown in Figure 3. A detail of a wall of the burial mound is given in Figure 4.

Methods
Brick samples were used for testing purposes. In detail, samples were subjected to visual analysis, water absorption and weight density measurements, bending and compressive testing, and mineralogical analysis. Table 1 shows the test matrix. The reader should be alerted to the limited experimental data base (also considering that the test results lack the statistical properties) from which these analyses started. In order to obtain prior information about the properties of the brick samples taken from the historical structures, visual analysis of the brick samples was performed. Initially, inconsistent material on the surfaces of the brick samples was removed by means of an air compressor. After cleaning their surfaces, brick samples were examined thoroughly. It is common to find natural fibers embedded into the structure of solid bricks from this period and region. However, this was not the case in the bricks extracted in Harran: there were no natural fibers (organic additive) present in the tested samples (Table 1). In order to determine the original colors of the brick samples, which surface dusts, the colors of the brick samples were determined using a colorimeter (model X-Rite RM200), and the color codes found out were specified by the corresponding colors on the color chart.

Physical Analyses
The test for water absorption is generally a good indication of the freeze/thaw resistance, and it provides important information about the durability of bricks. Water absorption is measured by how much a brick increases in weight after it has been soaked in water. In order to determine the water absorption rates of the bricks used at Harran, a series of laboratory tests were conducted [46].

Methods
Brick samples were used for testing purposes. In detail, samples were subjected to visual analysis, water absorption and weight density measurements, bending and compressive testing, and mineralogical analysis. Table 1 shows the test matrix. The reader should be alerted to the limited experimental data base (also considering that the test results lack the statistical properties) from which these analyses started.

Visual Analyses
In order to obtain prior information about the properties of the brick samples taken from the historical structures, visual analysis of the brick samples was performed. Initially, inconsistent material on the surfaces of the brick samples was removed by means of an air compressor. After cleaning their surfaces, brick samples were examined thoroughly. It is common to find natural fibers embedded into the structure of solid bricks from this period and region. However, this was not the case in the bricks extracted in Harran: there were no natural fibers (organic additive) present in the tested samples (Table 1). In order to determine the original colors of the brick samples, which surface dusts, the colors of the brick samples were determined using a colorimeter (model X-Rite RM200), and the color codes found out were specified by the corresponding colors on the color chart.

Physical Analyses
The test for water absorption is generally a good indication of the freeze/thaw resistance, and it provides important information about the durability of bricks. Water absorption is measured by how much a brick increases in weight after it has been soaked in water. In order to determine the water absorption rates of the bricks used at Harran, a series of laboratory tests were conducted [46]. Initially inconsistent material, traces of organic and inorganic materials (lime mortar, mold, etc.) had been removed from the brick samples via an air compressor. Subsequently, the brick samples were dried in an oven for 48 h at 60 • C and cooled in a desiccator. The samples were dried until the weight gain remained unchanged. At the end of this procedure, the samples' dry weights were recorded. Then, brick specimens were soaked for 48 hours in water, after which their water-logged and underwater weights were recorded.
The aim of this investigation was also to calculate the density of the bricks used at the site of Harran. It is well-known that density is an elementary physical property, and, for a homogeneous object, it is defined as the ratio of its mass (m) to its volume (V). However, due to the amorphous form of the brick samples, apparent density was initially calculated using the results of the water absorption tests [47,48], by considering the volume and weights of the brick specimens, in their dry and water-impregnated states.
However, pycnometer tests were also conducted in order to measure the real densities of the brick samples [49]. Furthermore, to calculate the porosity values of the brick samples, compactness was calculated from the test results in terms of the apparent and the real densities [50].

Mechanical and Mineralogical Analyses
Samples were prepared by reducing the bricks to the dimensions of 40 × 40 × 160 mm with a hacksaw. Prepared samples were placed on the testing machine on one of the lateral surfaces upon the support cylinders in such a way that the longitudinal axis would be normal to the support cylinders ( Figure 5). The load was applied normal to the lateral surface of the prism via the loader cylinder, with its axis normal to the longitudinal axis of the specimen, and the load was increased gradually at the rate of (50 ± 10) N/s until the sample was broken [51].
Heritage 2019, 3 FOR PEER REVIEW 7 Initially inconsistent material, traces of organic and inorganic materials (lime mortar, mold, etc.) had been removed from the brick samples via an air compressor. Subsequently, the brick samples were dried in an oven for 48 h at 60 °C and cooled in a desiccator. The samples were dried until the weight gain remained unchanged. At the end of this procedure, the samples' dry weights were recorded. Then, brick specimens were soaked for 48 hours in water, after which their water-logged and underwater weights were recorded. The aim of this investigation was also to calculate the density of the bricks used at the site of Harran. It is well-known that density is an elementary physical property, and, for a homogeneous object, it is defined as the ratio of its mass (m) to its volume (V). However, due to the amorphous form of the brick samples, apparent density was initially calculated using the results of the water absorption tests [47,48], by considering the volume and weights of the brick specimens, in their dry and water-impregnated states.
However, pycnometer tests were also conducted in order to measure the real densities of the brick samples [49]. Furthermore, to calculate the porosity values of the brick samples, compactness was calculated from the test results in terms of the apparent and the real densities [50].

Mechanical and Mineralogical Analyses
Samples were prepared by reducing the bricks to the dimensions of 40 × 40 × 160 mm with a hacksaw. Prepared samples were placed on the testing machine on one of the lateral surfaces upon the support cylinders in such a way that the longitudinal axis would be normal to the support cylinders ( Figure 5). The load was applied normal to the lateral surface of the prism via the loader cylinder, with its axis normal to the longitudinal axis of the specimen, and the load was increased gradually at the rate of (50 ± 10) N/s until the sample was broken [51]. The compressive strengths of the brick samples were determined by measuring the uniaxial compressive strength, applying the compressive load on the prismatic samples obtained from the broken halves of the bending strength test specimens [52]. The bending strength (sbend) was initially calculated using: where Fsd is the maximum bending load, L0 is the bending span (100 mm), b and d are the width and height of the brick specimen (40 mm). And the compressive strength (scompr): where Fsc is the maximum compressive load. In this analysis, which helps the determination of the texture and minerals of brick samples, at first the samples in bulk, which would represent the petrographic properties of the material, were dried for 2 h under 105 ± 5 °C after the dusts on them had been brushed away. The dried samples The compressive strengths of the brick samples were determined by measuring the uniaxial compressive strength, applying the compressive load on the prismatic samples obtained from the broken halves of the bending strength test specimens [52]. The bending strength (s bend ) was initially calculated using: where F sd is the maximum bending load, L 0 is the bending span (100 mm), b and d are the width and height of the brick specimen (40 mm). And the compressive strength (s compr ): where F sc is the maximum compressive load. In this analysis, which helps the determination of the texture and minerals of brick samples, at first the samples in bulk, which would represent the petrographic properties of the material, were dried for 2 h under 105 ± 5 • C after the dusts on them had been brushed away. The dried samples were first let to absorb epoxy polymer under pressure and let to dry. Then, they were cut in thin sections in a section-preparation device and were glued with epoxy polymer. After the necessary cutting and thinning processes in the thin-section device, they were polished. The texture of the sample, bounding aggregate ratios, bounding aggregate phases, aggregate types, shapes and sizes in the prepared polished sections were determined via a stereo microscope (single nicol). Later, these sections went through thinning and polishing processes so that they could be examined under the polarizing microscope. Working on the section prepared by thinning until the thickness of 30 µ was reached, the definite identification of minerals, decomposition, newly created minerals and the textural properties formed by them were examined by using the polarizing microscope (two nicols) [53].

Visual Analyses
The colors of bonding agent and aggregates, quantity and type of aggregates within brick, and, if present, organic additives within brick ingredients were determined by qualitative observations ( Table 2). were first let to absorb epoxy polymer under pressure and let to dry. Then, they were cut in thin sections in a section-preparation device and were glued with epoxy polymer. After the necessary cutting and thinning processes in the thin-section device, they were polished. The texture of the sample, bounding aggregate ratios, bounding aggregate phases, aggregate types, shapes and sizes in the prepared polished sections were determined via a stereo microscope (single nicol). Later, these sections went through thinning and polishing processes so that they could be examined under the polarizing microscope. Working on the section prepared by thinning until the thickness of 30 μ was reached, the definite identification of minerals, decomposition, newly created minerals and the textural properties formed by them were examined by using the polarizing microscope (two nicols) [53].

Visual Analyses
The colors of bonding agent and aggregates, quantity and type of aggregates within brick, and, if present, organic additives within brick ingredients were determined by qualitative observations ( Table 2). The specific colors of the brick samples are close to those of the clays contained therein. In terms of strength, the brick samples are strong enough such that they cannot be broken by hand. The shapes of the brick samples are rectangular and square. Rectangular and square shapes enable easier and regular building of walls.

Physical Analyses
Physical analyses allowed us to determine the water absorption rates by weight and by volume of the brick samples. Apart from the compressive strength, it is well-known that water absorption is an important property in brickwork masonry. This property can affect the quality of the brick itself and the bond strength between the brick and mortar in masonry structures and can result in reducing its strength properties. The authors [25] have previously conducted an experimental investigation to assess the physical, mineralogical, and chemical properties of the lime mortars used in the monuments of Harran. The test results carried out on the lime mortar and on the units (i.e., the bricks) could be critical for any future conservation measure and assessment of the structural safety of the masonry structures of the site of Harran: this could highly facilitate the estimation of the mechanical properties of the brickwork masonry. Water absorption rates by weight and by volume of the samples are given in Table 3. Figure 6 shows the tested specimen during soaking in water.

Walls
No HU Heritage 2019, 3 FOR PEER REVIEW 8 were first let to absorb epoxy polymer under pressure and let to dry. Then, they were cut in thin sections in a section-preparation device and were glued with epoxy polymer. After the necessary cutting and thinning processes in the thin-section device, they were polished. The texture of the sample, bounding aggregate ratios, bounding aggregate phases, aggregate types, shapes and sizes in the prepared polished sections were determined via a stereo microscope (single nicol). Later, these sections went through thinning and polishing processes so that they could be examined under the polarizing microscope. Working on the section prepared by thinning until the thickness of 30 μ was reached, the definite identification of minerals, decomposition, newly created minerals and the textural properties formed by them were examined by using the polarizing microscope (two nicols) [53].

Visual Analyses
The colors of bonding agent and aggregates, quantity and type of aggregates within brick, and, if present, organic additives within brick ingredients were determined by qualitative observations ( Table 2). The specific colors of the brick samples are close to those of the clays contained therein. In terms of strength, the brick samples are strong enough such that they cannot be broken by hand. The shapes of the brick samples are rectangular and square. Rectangular and square shapes enable easier and regular building of walls.

Physical Analyses
Physical analyses allowed us to determine the water absorption rates by weight and by volume of the brick samples. Apart from the compressive strength, it is well-known that water absorption is an important property in brickwork masonry. This property can affect the quality of the brick itself and the bond strength between the brick and mortar in masonry structures and can result in reducing its strength properties. The authors [25] have previously conducted an experimental investigation to assess the physical, mineralogical, and chemical properties of the lime mortars used in the monuments of Harran. The test results carried out on the lime mortar and on the units (i.e., the bricks) could be critical for any future conservation measure and assessment of the structural safety of the masonry structures of the site of Harran: this could highly facilitate the estimation of the mechanical properties of the brickwork masonry. Water absorption rates by weight and by volume of the samples are given in Table 3. Figure 6 shows the tested specimen during soaking in water.

Partition walls No
HCW Heritage 2019, 3 FOR PEER REVIEW 8 were first let to absorb epoxy polymer under pressure and let to dry. Then, they were cut in thin sections in a section-preparation device and were glued with epoxy polymer. After the necessary cutting and thinning processes in the thin-section device, they were polished. The texture of the sample, bounding aggregate ratios, bounding aggregate phases, aggregate types, shapes and sizes in the prepared polished sections were determined via a stereo microscope (single nicol). Later, these sections went through thinning and polishing processes so that they could be examined under the polarizing microscope. Working on the section prepared by thinning until the thickness of 30 μ was reached, the definite identification of minerals, decomposition, newly created minerals and the textural properties formed by them were examined by using the polarizing microscope (two nicols) [53].

Visual Analyses
The colors of bonding agent and aggregates, quantity and type of aggregates within brick, and, if present, organic additives within brick ingredients were determined by qualitative observations ( Table 2). The specific colors of the brick samples are close to those of the clays contained therein. In terms of strength, the brick samples are strong enough such that they cannot be broken by hand. The shapes of the brick samples are rectangular and square. Rectangular and square shapes enable easier and regular building of walls.

Physical Analyses
Physical analyses allowed us to determine the water absorption rates by weight and by volume of the brick samples. Apart from the compressive strength, it is well-known that water absorption is an important property in brickwork masonry. This property can affect the quality of the brick itself and the bond strength between the brick and mortar in masonry structures and can result in reducing its strength properties. The authors [25] have previously conducted an experimental investigation to assess the physical, mineralogical, and chemical properties of the lime mortars used in the monuments of Harran. The test results carried out on the lime mortar and on the units (i.e., the bricks) could be critical for any future conservation measure and assessment of the structural safety of the masonry structures of the site of Harran: this could highly facilitate the estimation of the mechanical properties of the brickwork masonry. Water absorption rates by weight and by volume of the samples are given in Table 3. Figure 6 shows the tested specimen during soaking in water.

Partition walls No
HC Heritage 2019, 3 FOR PEER REVIEW 8 were first let to absorb epoxy polymer under pressure and let to dry. Then, they were cut in thin sections in a section-preparation device and were glued with epoxy polymer. After the necessary cutting and thinning processes in the thin-section device, they were polished. The texture of the sample, bounding aggregate ratios, bounding aggregate phases, aggregate types, shapes and sizes in the prepared polished sections were determined via a stereo microscope (single nicol). Later, these sections went through thinning and polishing processes so that they could be examined under the polarizing microscope. Working on the section prepared by thinning until the thickness of 30 μ was reached, the definite identification of minerals, decomposition, newly created minerals and the textural properties formed by them were examined by using the polarizing microscope (two nicols) [53].

Visual Analyses
The colors of bonding agent and aggregates, quantity and type of aggregates within brick, and, if present, organic additives within brick ingredients were determined by qualitative observations ( Table 2). The specific colors of the brick samples are close to those of the clays contained therein. In terms of strength, the brick samples are strong enough such that they cannot be broken by hand. The shapes of the brick samples are rectangular and square. Rectangular and square shapes enable easier and regular building of walls.

Physical Analyses
Physical analyses allowed us to determine the water absorption rates by weight and by volume of the brick samples. Apart from the compressive strength, it is well-known that water absorption is an important property in brickwork masonry. This property can affect the quality of the brick itself and the bond strength between the brick and mortar in masonry structures and can result in reducing its strength properties. The authors [25] have previously conducted an experimental investigation to assess the physical, mineralogical, and chemical properties of the lime mortars used in the monuments of Harran. The test results carried out on the lime mortar and on the units (i.e., the bricks) could be critical for any future conservation measure and assessment of the structural safety of the masonry structures of the site of Harran: this could highly facilitate the estimation of the mechanical properties of the brickwork masonry. Water absorption rates by weight and by volume of the samples are given in Table 3. Figure 6 shows the tested specimen during soaking in water.

Partition walls No
HGM Heritage 2019, 3 FOR PEER REVIEW 8 were first let to absorb epoxy polymer under pressure and let to dry. Then, they were cut in thin sections in a section-preparation device and were glued with epoxy polymer. After the necessary cutting and thinning processes in the thin-section device, they were polished. The texture of the sample, bounding aggregate ratios, bounding aggregate phases, aggregate types, shapes and sizes in the prepared polished sections were determined via a stereo microscope (single nicol). Later, these sections went through thinning and polishing processes so that they could be examined under the polarizing microscope. Working on the section prepared by thinning until the thickness of 30 μ was reached, the definite identification of minerals, decomposition, newly created minerals and the textural properties formed by them were examined by using the polarizing microscope (two nicols) [53].

Visual Analyses
The colors of bonding agent and aggregates, quantity and type of aggregates within brick, and, if present, organic additives within brick ingredients were determined by qualitative observations ( Table 2). The specific colors of the brick samples are close to those of the clays contained therein. In terms of strength, the brick samples are strong enough such that they cannot be broken by hand. The shapes of the brick samples are rectangular and square. Rectangular and square shapes enable easier and regular building of walls.

Physical Analyses
Physical analyses allowed us to determine the water absorption rates by weight and by volume of the brick samples. Apart from the compressive strength, it is well-known that water absorption is an important property in brickwork masonry. This property can affect the quality of the brick itself and the bond strength between the brick and mortar in masonry structures and can result in reducing its strength properties. The authors [25] have previously conducted an experimental investigation to assess the physical, mineralogical, and chemical properties of the lime mortars used in the monuments of Harran. The test results carried out on the lime mortar and on the units (i.e., the bricks) could be critical for any future conservation measure and assessment of the structural safety of the masonry structures of the site of Harran: this could highly facilitate the estimation of the mechanical properties of the brickwork masonry. Water absorption rates by weight and by volume of the samples are given in Table 3. Figure 6 shows the tested specimen during soaking in water.

Partitions of the mosque No
The specific colors of the brick samples are close to those of the clays contained therein. In terms of strength, the brick samples are strong enough such that they cannot be broken by hand. The shapes of the brick samples are rectangular and square. Rectangular and square shapes enable easier and regular building of walls.

Physical Analyses
Physical analyses allowed us to determine the water absorption rates by weight and by volume of the brick samples. Apart from the compressive strength, it is well-known that water absorption is an important property in brickwork masonry. This property can affect the quality of the brick itself and the bond strength between the brick and mortar in masonry structures and can result in reducing its strength properties. The authors [25] have previously conducted an experimental investigation to assess the physical, mineralogical, and chemical properties of the lime mortars used in the monuments of Harran. The test results carried out on the lime mortar and on the units (i.e., the bricks) could be critical for any future conservation measure and assessment of the structural safety of the masonry structures of the site of Harran: this could highly facilitate the estimation of the mechanical properties of the brickwork masonry. Water absorption rates by weight and by volume of the samples are given in Table 3. Figure 6 shows the tested specimen during soaking in water. The water absorption rates of the brick samples by weight vary between 17.30% and 38.12%. The lowest water absorption rate by weight was obtained from brick samples taken from the city walls. The highest water absorption rate by weight was obtained from brick samples taken from the great mosque. The water absorption rates by volume of the brick samples vary between 25.54% and 45.43%. Brick samples taken from the castle have the lowest water absorption rate by volume. A mortar sample taken from the great mosque has the highest water absorption rate by volume. When the water absorption rates of the brick samples are evaluated in accordance with Turkish Standard 704 [54], it is seen that the water absorption rates are high. The reason for this might be the method or the ingredients used during their production. Furthermore, the volumes of the brick samples have been determined via water absorption test since they were not prismatic in shape. Then, apparent densities of the brick samples have been compiled (Table 4). The real densities of the brick samples vary between 1.33-1.70 g/cm 3 . The brick sample with the highest real density value is the one taken from the burial mound. The brick sample with the lowest real density value is the one taken from the great mosque. Real density values of the brick samples are within the range specified by Turkish Standard 704 (1.20-2.00 g/cm 3 ) [54]. Real densities of the brick samples falling within the range specified in the Turkish Standard indicates that a serious controlled production procedure has been used for making them. The water absorption rates of the brick samples by weight vary between 17.30% and 38.12%. The lowest water absorption rate by weight was obtained from brick samples taken from the city walls. The highest water absorption rate by weight was obtained from brick samples taken from the great mosque. The water absorption rates by volume of the brick samples vary between 25.54% and 45.43%. Brick samples taken from the castle have the lowest water absorption rate by volume. A mortar sample taken from the great mosque has the highest water absorption rate by volume. When the water absorption rates of the brick samples are evaluated in accordance with Turkish Standard 704 [54], it is seen that the water absorption rates are high. The reason for this might be the method or the ingredients used during their production.
Furthermore, the volumes of the brick samples have been determined via water absorption test since they were not prismatic in shape. Then, apparent densities of the brick samples have been compiled (Table 4). The real densities of the brick samples vary between 1.33-1.70 g/cm 3 . The brick sample with the highest real density value is the one taken from the burial mound. The brick sample with the lowest real density value is the one taken from the great mosque. Real density values of the brick samples are within the range specified by Turkish Standard 704 (1.20-2.00 g/cm 3 ) [54]. Real densities of the brick samples falling within the range specified in the Turkish Standard indicates that a serious controlled production procedure has been used for making them.
Porosity values of the brick samples were calculated using their apparent densities and real densities. During the firing of the bricks, infinite microscopic cracks, cavities, and gaps are formed (pores) [55]. These pores are, especially, effective in the failures of the bricks. In general, when the porosity of a brick goes above 25%, its compressive strength gradually decreases. One of the most significant factors which affected the durability of the brick samples is their porosity. Since harmful liquids easily penetrate into the bricks with higher porosity values, their durability decreases. The porosity values of the brick samples vary between 8.88% and 25.31%. The brick sample with the highest porosity value is the one taken from the burial mound. The sample with the lowest porosity value is the one taken from the city walls. Since porosity value affects the strength and durability of a brick, the porosity must be kept below a particular value so that a desired brick quality is guaranteed. As the porosity values of the bricks studied in this research are low, that must be the reason why these historical structures have survived to date.
Compressive strength tests on the fired brick samples were carried out to determine the load-carrying capacity of bricks, together with analysis of the lime mortar [25] to estimate the compressive strength of the brickwork masonry. Because Turkey is one of the most active seismic regions in Europe and Asia (Figure 7), it was decided to measure the bending strength of the bricks. This mechanical property could be useful for an estimation of the tensile and shear strengths of the brickwork masonry used at the monuments of Harran, this providing interesting information about the seismic vulnerability of Harran buildings. Unfortunately, given the high level of protection by the Turkish authority over the site of Harran, it was not possible to conduct more exhaustive or full-scale testing on the brickwork masonry. The bending and compressive strengths of the brick samples are given in Table 5.
Heritage 2019, 3 FOR PEER REVIEW 10 Porosity values of the brick samples were calculated using their apparent densities and real densities. During the firing of the bricks, infinite microscopic cracks, cavities, and gaps are formed (pores) [55]. These pores are, especially, effective in the failures of the bricks. In general, when the porosity of a brick goes above 25%, its compressive strength gradually decreases. One of the most significant factors which affected the durability of the brick samples is their porosity. Since harmful liquids easily penetrate into the bricks with higher porosity values, their durability decreases. The porosity values of the brick samples vary between 8.88% and 25.31%. The brick sample with the highest porosity value is the one taken from the burial mound. The sample with the lowest porosity value is the one taken from the city walls. Since porosity value affects the strength and durability of a brick, the porosity must be kept below a particular value so that a desired brick quality is guaranteed. As the porosity values of the bricks studied in this research are low, that must be the reason why these historical structures have survived to date.
Compressive strength tests on the fired brick samples were carried out to determine the loadcarrying capacity of bricks, together with analysis of the lime mortar [25] to estimate the compressive strength of the brickwork masonry. Because Turkey is one of the most active seismic regions in Europe and Asia (Figure 7), it was decided to measure the bending strength of the bricks. This mechanical property could be useful for an estimation of the tensile and shear strengths of the brickwork masonry used at the monuments of Harran, this providing interesting information about the seismic vulnerability of Harran buildings. Unfortunately, given the high level of protection by the Turkish authority over the site of Harran, it was not possible to conduct more exhaustive or fullscale testing on the brickwork masonry. The bending and compressive strengths of the brick samples are given in Table 5. A large amount of information is available about the geology of the region of Harran. This is located in the basin of the Balikh river. This joins the Euphrates at the modern city of Raqqa. Figure  8 shows a geology map of the area of Harran: it can be noted that the rocks of the valley belong to the Quaternary period, being the most recently laid geologic strata. Based on the low seismic risk of the area of Harran and its geology, with likely a low local amplification over a rocky soil, we can conclude that the seismic hazard for the site of Harran is relatively low. However, the site deserves a more in-depth analysis, also considering hazards from climate change (heavy rain falls, flooding, etc.) and their combined effects on the masonry monuments of Harran. , in terms of the ratio between peak ground acceleration a g and g (g = acceleration of gravity), for a probability of exceedance of 10% in 50 years (475 years return period). A large amount of information is available about the geology of the region of Harran. This is located in the basin of the Balikh river. This joins the Euphrates at the modern city of Raqqa. Figure 8 shows a geology map of the area of Harran: it can be noted that the rocks of the valley belong to the Quaternary period, being the most recently laid geologic strata. Based on the low seismic risk of the area of Harran and its geology, with likely a low local amplification over a rocky soil, we can conclude that the seismic hazard for the site of Harran is relatively low. However, the site deserves a more in-depth analysis, also considering hazards from climate change (heavy rain falls, flooding, etc.) and their combined effects on the masonry monuments of Harran.   Figure 9 shows several brick specimens before these were reduced to the standard testing dimensions. This figure also shows the testing instrument used for both bending and compression tests. The bending strengths bend of the brick samples vary between 0.82 and 1.86 MPa. The brick sample with the highest bending strength is the one taken from the burial mound. The brick sample with the lowest bending strength is the one taken from the great mosque. The compressive strengths compr of the brick samples vary between 6.69 and 7.95 MPa. The brick sample with the highest compressive strength value is the one taken from the burial mound. Figure 10 shows a brick specimen before being subjected to the compressive load: the underlying brick debris from previous tests clearly demonstrate that the samples' failure mode was due to crumbling/pulverization of the tile material. This is a well-known failure mechanism that is normal for low-strength, porous masonry materials [57].  Figure 9 shows several brick specimens before these were reduced to the standard testing dimensions. This figure also shows the testing instrument used for both bending and compression tests. The bending strengths σ bend of the brick samples vary between 0.82 and 1.86 MPa. The brick sample with the highest bending strength is the one taken from the burial mound. The brick sample with the lowest bending strength is the one taken from the great mosque.   Figure 9 shows several brick specimens before these were reduced to the standard testing dimensions. This figure also shows the testing instrument used for both bending and compression tests. The bending strengths bend of the brick samples vary between 0.82 and 1.86 MPa. The brick sample with the highest bending strength is the one taken from the burial mound. The brick sample with the lowest bending strength is the one taken from the great mosque. The compressive strengths compr of the brick samples vary between 6.69 and 7.95 MPa. The brick sample with the highest compressive strength value is the one taken from the burial mound. Figure 10 shows a brick specimen before being subjected to the compressive load: the underlying brick debris from previous tests clearly demonstrate that the samples' failure mode was due to crumbling/pulverization of the tile material. This is a well-known failure mechanism that is normal for low-strength, porous masonry materials [57]. The compressive strengths σ compr of the brick samples vary between 6.69 and 7.95 MPa. The brick sample with the highest compressive strength value is the one taken from the burial mound. Figure 10 shows a brick specimen before being subjected to the compressive load: the underlying brick debris from previous tests clearly demonstrate that the samples' failure mode was due to crumbling/pulverization of the tile material. This is a well-known failure mechanism that is normal for low-strength, porous masonry materials [57]. The sample with the lowest compressive strength value is the one taken from the city walls. According to previous studies [58], it is possible to say that the compressive strength values of the brick samples are very similar and in a limited range (6.69-7.95 MPa). This could suggest that the Harran monuments, constructed in different periods of time, have been manufactured using a similar or non-time-varying technology, labor, resources, and environmental conditions. This could strengthen the hypothesis that local skills, construction materials, and methods were passed down from father to son for generations in family brick companies, also considering the frequent changing of rulers (from Parthians to Romans, from Byzantines to Arabs, and finally to the Mongols and Turkish). The thin-section images of the five brick samples taken from the historical structures are quite similar to each other ( Figure 11). In order to scan the minerals and the pores in the brick samples more clearly, single-and two-Nicol prism filters were used to capture the images. There is an excessive amount of calcite mineral in the HBM brick sample. Also, it can be noted that it contains muscovite, plagioclase, and oxide minerals. These minerals are bonded with calcite ( Figure 12). Excessive amounts of calcite mineral are also present in the HU brick sample. It was also concluded that it has a porous structure, The sample with the lowest compressive strength value is the one taken from the city walls. According to previous studies [58], it is possible to say that the compressive strength values of the brick samples are very similar and in a limited range (6.69-7.95 MPa). This could suggest that the Harran monuments, constructed in different periods of time, have been manufactured using a similar or non-time-varying technology, labor, resources, and environmental conditions. This could strengthen the hypothesis that local skills, construction materials, and methods were passed down from father to son for generations in family brick companies, also considering the frequent changing of rulers (from Parthians to Romans, from Byzantines to Arabs, and finally to the Mongols and Turkish). The thin-section images of the five brick samples taken from the historical structures are quite similar to each other ( Figure 11). The sample with the lowest compressive strength value is the one taken from the city walls. According to previous studies [58], it is possible to say that the compressive strength values of the brick samples are very similar and in a limited range (6.69-7.95 MPa). This could suggest that the Harran monuments, constructed in different periods of time, have been manufactured using a similar or non-time-varying technology, labor, resources, and environmental conditions. This could strengthen the hypothesis that local skills, construction materials, and methods were passed down from father to son for generations in family brick companies, also considering the frequent changing of rulers (from Parthians to Romans, from Byzantines to Arabs, and finally to the Mongols and Turkish). The thin-section images of the five brick samples taken from the historical structures are quite similar to each other ( Figure 11). In order to scan the minerals and the pores in the brick samples more clearly, single-and two-Nicol prism filters were used to capture the images. There is an excessive amount of calcite mineral in the HBM brick sample. Also, it can be noted that it contains muscovite, plagioclase, and oxide minerals. These minerals are bonded with calcite ( Figure 12). Excessive amounts of calcite mineral are also present in the HU brick sample. It was also concluded that it has a porous structure, Figure 11. Detail of the brick cross-section.
In order to scan the minerals and the pores in the brick samples more clearly, single-and two-Nicol prism filters were used to capture the images. There is an excessive amount of calcite mineral in the HBM brick sample. Also, it can be noted that it contains muscovite, plagioclase, and oxide minerals. These minerals are bonded with calcite ( Figure 12). Excessive amounts of calcite mineral are also present in the HU brick sample. It was also concluded that it has a porous structure, containing, in addition to calcite minerals, plagioclase and quartz minerals. Plenty of quartz and calcite minerals are present in the HCW brick sample, consisting of tiny grains.
Heritage 2019, 3 FOR PEER REVIEW 13 containing, in addition to calcite minerals, plagioclase and quartz minerals. Plenty of quartz and calcite minerals are present in the HCW brick sample, consisting of tiny grains. It could be interesting to compare the brick mechanical properties of the Harran monuments with the ones of other sites in the region. Because most of these sites are listed and protected it is very rare to find these values in the scientific literature, given the destructive character of these tests. Table 6 shows these data for a limited number of sites located in Greece and Turkey. It can be noted that the compressive strengths of Harran bricks are consistent with the mechanical characteristics of the Serapis temple of Pergamon (second century), in Bergama, Turkey, and with the ones of Ottoman Ö rdekli Bath, Bursa, Turkey, but smaller compared to the compressive strengths of some Roman and Byzantine masonry monuments located in Filippi, Greece.
On the other hand, the density of Harran bricks seems significantly smaller compared to the density of bricks used in Turkey in historic brickwork constructions. For example, the porosity of the bricks of the Serapis temple, Pergamon; Agia Sophia, Istanbul; and Ottoman Ö rdekli Bath, Bursa, are 35, 45, and 39%, respectively (Table 5). Harran bricks exhibited a porosity ranging between 8.88 and 25.31 %, with a mean value of about 13% (Table 4). By comparing the mechanical properties of the bricks and the mortar used in Harran with other archeological monuments in the macro-Mediterranean region, critical information can be obtained. For example, conservation interventions successfully implemented in other sites, based on their property similarity with the masonry materials, can be adopted with more confidence. Conservation of archeological monuments requires a deep knowledge of structures and constituent It could be interesting to compare the brick mechanical properties of the Harran monuments with the ones of other sites in the region. Because most of these sites are listed and protected it is very rare to find these values in the scientific literature, given the destructive character of these tests. Table 6 shows these data for a limited number of sites located in Greece and Turkey. It can be noted that the compressive strengths of Harran bricks are consistent with the mechanical characteristics of the Serapis temple of Pergamon (second century), in Bergama, Turkey, and with the ones of Ottoman Ördekli Bath, Bursa, Turkey, but smaller compared to the compressive strengths of some Roman and Byzantine masonry monuments located in Filippi, Greece. On the other hand, the density of Harran bricks seems significantly smaller compared to the density of bricks used in Turkey in historic brickwork constructions. For example, the porosity of the bricks of the Serapis temple, Pergamon; Agia Sophia, Istanbul; and Ottoman Ördekli Bath, Bursa, are 35, 45, and 39%, respectively (Table 5). Harran bricks exhibited a porosity ranging between 8.88 and 25.31 %, with a mean value of about 13% (Table 4).
By comparing the mechanical properties of the bricks and the mortar used in Harran with other archeological monuments in the macro-Mediterranean region, critical information can be obtained. For example, conservation interventions successfully implemented in other sites, based on their property similarity with the masonry materials, can be adopted with more confidence. Conservation of archeological monuments requires a deep knowledge of structures and constituent materials, of their properties, context features and use conditions, and of the possible state of damage and its causes. Prevention and rehabilitation measures can be successfully accomplished only if a proper analysis and diagnosis has been formulated, so that masonry heritage buildings can be preserved as much as possible as historic documents of our past.
For the Harran site, several questions remain without answers: this relates to the sites from where the inhabitants of Harran collected the raw materials used for brick production, the technologies they applied to make and fire the bricks. It could be very interesting to study how far they transported raw resources and final products after furnace treatments, considering that a crucial point was the nearby availability of timber, water, and sandy soils without stones. This paper aims to provide a limited, but significant contribution to these critical issues.

Conclusions
In this paper the physical-mechanical and mineralogical properties of bricks used in historical masonry structures of the ancient city of Harran were investigated. At the end of the study, the following conclusions can be drawn:

1.
The visual analyses of the bricks allowed to describe them based on their colors. Importantly, this is a potential source of cultural and heritage information and can provide useful information about the identity, chronology, technology, labor, resources, and environmental conditions of the site of Harran, in Turkey.

2.
The mechanical properties (compressive and bending strengths), porosity, and density of the bricks have been compared with the scarce existing data for other sites in the macro-region (Asia minor, southern Europe, northern Africa). Also taking into account the limited experimental base, it can be concluded that the mechanical properties are consistent with the ones of other sites, even if a large scattering of results was noted from one site to another. It seems that bricks' mechanical characteristics from Roman times for sites in Greece are higher (18-20 MPa for the compressive strength) compared with the compressive strength of the buildings constructed in the Late Antiquity (3rd-10th century).

3.
It is also possible to note that the compressive and bending strength values of the tested brick samples are very similar and in a limited range (6.69-7.95 MPa for compressive strength, and 0.82-1.86 MPa for bending strength). This could suggest that the Harran masonry monuments, constructed in different periods of time, have been manufactured using a similar or a non-time-varying technology, labor, resources, and environmental conditions. This could strengthen the hypothesis that local skills, construction materials, and methods were passed down from father to son for generations in the family brick companies and were not shattered by the destructions and devastation from frequent wars, changing of rulers, natural disasters that affected the area of Harran. 4.
The mechanical characterization of the masonry constituent materials (bricks and mortar) of several Harran monuments represents a critical information for future analyses regarding the seismic vulnerability, and the implementation of conservation interventions.

5.
The minerals seen in the thin-section images of the bricks are cryptocrystals, quartz, ferro magnesium, calcite, and opaque minerals. It can be said that the ferro magnesium minerals within the bricks provide a binding material support to the bricks by decomposition.
These structures in Harran Area, which were built centuries ago (since 5000 BC), provided great services. These structures are extremely valuable as cultural heritage. Thus, before it is too late, they need to be restored and put under protection. With planned restorations on these structures, by taking into consideration the scientific studies conducted on them, being our cultural heritage, they must be handed on to the coming generations. Funding: This research received no external funding.

Conflicts of Interest:
The authors declare no conflict of interest.