Characterisation of ancient mortars from the S. Niccolò archaeological complex in Montieri (Tuscany – Italy)

Construction and Building Materials 96 (2015) 442–460

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Characterisation of ancient mortars from the S. Niccolò archaeological complex in Montieri (Tuscany – Italy) Nicoletta Chiarelli a,⇑, Domenico Miriello b, Giovanna Bianchi c, Giuseppe Fichera c, Marco Giamello a, Isabella Turbanti Memmi a a b c

Department of Physical Sciences, Earth and Environment, University of Siena, Italy Department of Biology, Ecology and Earth Sciences, University of Calabria, Italy Department of Historical Sciences and Cultural Heritage, University of Siena, Italy

h i g h l i g h t s  The study of the mortars allowed us to identify 5 principal building phases.  Metallurgic slags, cocciopesto and glasses are visible in the mortar aggregate.  The metallurgic slags give to the binder a greater hydraulicity.  Argentiferous minerals are present in the aggregate of some mortars.

a r t i c l e

i n f o

Article history: Received 10 March 2015 Received in revised form 17 July 2015 Accepted 5 August 2015

Keywords: Archaeometry OM XRD SEM–EDS XRF Lime Lump Slag

a b s t r a c t The object of the present work is the archaeometric study of the mortars from the S. Niccolò archaeological site in Montieri (Tuscany – Italy). The petrographic, mineralogical and chemical characterisation of the samples was obtained by the use of multi-analytical techniques (OM, XRD, SEM–EDS, XRF). The knowledge of the constitutive materials allowed us to understand the different production technologies and highlighted the use of materials with hydraulic behaviour, such as metallurgical slags and cocciopesto. The compositional analogies and differences among the samples contributed to the reconstruction of the building phases, favouring the reading of the history and the evolution of the site over time. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction Mortars have been employed for the edification of historical monuments since the ancient times. They can be defined as artificial stone materials, consisting of a mixture of inorganic or organic binders, aggregate, water and inorganic or organic additives. The proportions of the previous single components should give suitable workability and appropriate physical, mechanical and aesthetic characteristics to the finished product [1]. Today, a great number of researches focus the attention on the study of historical mortars, as source material of important information in archaeological, conservative and scientific fields [2–12]. The investigation about the ⇑ Corresponding author. E-mail address: [email protected] (N. Chiarelli). http://dx.doi.org/10.1016/j.conbuildmat.2015.08.023 0950-0618/Ó 2015 Elsevier Ltd. All rights reserved.

nature and the provenance of the raw materials, the productive technologies, the change of the production practices (lime burning, mortar mixing), the building phases and the dating of the builds are interesting data in historical and archaeological contexts, in order to reconstruct the economic and political scene of the countries. The compatible materials for restoration can be planned through the study of the original compositions and the deep knowledge of the deterioration processes. Furthermore the chemical, physical and mechanical processes in the mortars, such as the burning of the binder, hydration, carbonation and production of the hydraulic phases, are analysed for a better understanding of the behaviour and potentiality of these materials. The present work is focussed on the study of ancient mortars coming from the Canonica of S. Niccolò, a medieval archaeological site, located in the Colline Metallifere area (Tuscany)

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

443

Fig. 1. Location of the archaeological site, (a) map, (b) aerial photo and (c) wall rests of the church.

Fig. 2. Sampling, (a) location of the samples on the map, (b) macrophotography of the mortar MM1 and (c, d) sampling of the mortar MM19 and MM1.

(Fig. 1a and b). The development of the site is closely related to the near centre of Montieri (Grosseto), founded for the control and exploitation of the mines in the district. The aim of the study is the compositional characterization of the mortar samples, extensively taken from the wall remains of the archaeological site. For this purpose a multi-analytical approach is necessary, by using different analytical techniques:

the petrographic study in thin section of the samples by optical microscopy (OM); the mineralogical analysis by X-ray diffraction (XRD); micro-chemical investigations by scanning electron microscope (SEM–EDS) and bulk chemical analysis by X-ray fluorescence (XRF). The combination of this methodology with archaeological studies provides useful data to understand different aspects linked to the production technology of the mortars, the provenance of the

444

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

Table 1 Sampling data of the bedding mortars (MM: Montieri Mortar). d.g.l.: distance from the ground level, d.r.e.: distance from the right edge, d.l.e.: distance from the left edge, U.S.: stratigraphic unit, n.d.: not determined. Location

MM1 MM2 MM3 MM4 MM5 MM6 MM7 MM8 MM9 MM10 MM11 MM12 MM13 MM14 MM15 MM16 MM17 MM18 MM19 MM20 MM21 MM22 MM23 MM24 MM25 MM26 MM27 MM28 MM29 MM30 MM31 MM32 MM33 MM34 MM35 MM36 MM37 MM38 MM39 MM40 MM41 MM42 MM43 MM44 MM45 MM46 MM47 MM48 MM49 MM50 MM51 MM52 MM53 MM54

Area

d.g.l. (cm)

d.r.e. (cm)

d.l.e. (cm)

Area 1000 – 2nd building – foundation wall southeast apse Area 1000 – 2nd building – southeast apse Area 1000 – 2nd building – southeast apse Area 1000 – 2nd building – northeast apse Area 1000 – 2nd building – north apse Area 1000 – 2nd building – jamb of the north entrance Area 1000 – 2nd building – jamb of the north entrance Area 1000 – 2nd building – north apse Area 1000 – 2nd building – northwest apse Area 1000 – 2nd building – jamb of the south entrance Area 1000 – 2nd building – southeast apse Area 1000 – 2nd building – south apse Area 1000 – 2nd building – south apse Area 1000 – 1st building – U.S. 302 Area 1000 – 1st building – U.S. 203 Area 1000 – 1st building – foundation wall U.S. 206 Area 1000 – 1st building – semicircular wall Area 1000 – 1st building – semicircular wall Area 2000 – U.S. 1003 Area 2000 – U.S. 1003 Area 2000 – U.S. 1003 Area 2000 – U.S. 1001 Area 2000 – U.S. 1006 area 2000 - U.S. 803 Area 2000 – U.S. 801 Area 2000 – foundation wall U.S. 1106 Area 2000 – U.S. 1105 Area 2000 – U.S. 1101 Area 2000 Area 1000 – 1st building Area 4000 – U.S. 601 Area 4000 – U.S. 402 Area 4000 Area 4000 – U.S. 502 Outer wall of boundary – foundation wall U.S. 907 Outer wall of boundary – U.S. 906 Outer wall of boundary – U.S. 901 Area 4000 Area 9000 – jamb of entrance Area 9000 – jamb of entrance – U.S. 1205 Area 9000 Area 1000 – 1st building – floor level U.S. 302–303 Area 1000–1st building Area 1000 – 1st building – foundation wall U.S. 102 Area 1000 – 1st building – U.S. 101 Area 1000–1st building Area 1000 – 2nd building – foundation wall southwest apse Area 1000 – 2nd building – floor level southwest apse Area 1000–2nd building – southwest apse Area 1000–2nd building – jamb of the southwest entrance Area 1000 – 2nd building – southwest apse Area 1000 – 2nd building – foundation wall northeast apse area 1000–2nd building – northeast apse Area 1000–2nd building – northeast apse

17 58 86 50 39 42 32 30 54 101 82 85 100 35 110 80 80 67 58 13 24 50 110 30 65 17 35 100 20 20 40 90 90 95 26 100 120 50 0 75 50 21 0 70 80 42 25 45 100 50 48 60 90 100

92 130 152 60 – 4 0 57 160 12 73 15 60 – – – 125 160 160 65 60 – – 230 260 – – – – – – – – – 330 295 – – 0 0 – 145 – – 0 – – – – 250 248 120 270 180

– – – – 26 – – – – – – – – 120 190 197 – – – – – 120 230 – – 398 468 636 25 230 60 193 100 270 – – 75 150 – – 50 – 0 140 – 320 90 85 270 – – – – –

raw materials and the identification of the different building phases, in order to understand the growth and changes of the site in time. This approach becomes an useful tool, that can be used in other archaeological sites to solve archaeometric problems regarding ancient mortars and plasters. 1.1. Historical context and results of archaeological investigations The Colline Metallifere is an area with some of the richest silver deposits in Europe. Underground veins of minerals were mined to get raw materials, silver and copper, for the production of coins, from the early medieval period. The main political figures involved in mining these resources were members of important aristocratic families, monasteries of European standing and bishops.

Stratification

Adherence

– – – – – 2 layer – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – 2 layer – – – – – –

Poor Poor Poor Good Good Good Good Poor Good Very poor Good n.d. n.d. Good Poor Poor Poor Poor Good Good Good Moderate Poor Good Good Poor Moderate Good Poor Moderate Moderate Good Good Good Good Moderate Good Good n.d. Good Poor Good Good Poor Moderate Poor Poor Good Good Good n.d. n.d. Poor Poor

The richness of this evidence stimulated the interest of the archaeologist of the University of Siena and since the mid-80s, the ‘Progetto Colline Metallifere’ was launched. Connected to this is a large number of research projects, one of the main aims of which is to study relations between settlements and mineral resources, with particular reference to settlement dynamics and the dynamics of controlling the cycle of production in the medieval period. As of 2007, there began a large-scale project in the municipality of Montieri, including excavations at the Canonica [13,14]. The site is a few kilometres away from Montieri. The modern town corresponds to a small inhabited site whose size is not very different from the original medieval castle, first mentioned in 1133. Montieri was one of the main centres of silver production. The castle had links with the Bishop of Volterra, who, since the

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

445

Fig. 3. Photomicrographs of the mortar aggregates, (a) shale in sample MM54, (b) charcoal in sample MM11, (c) metallurgic slag in sample MM29, (d) glass in sample MM29, (e) opaque euhedral minerals in sample MM30, (f) iron oxides in sample MM27, (g) melilites in the metallurgic slag of the sample MM39 and (h) calcite in sample MM5.

end of the 12th century, set up a mint inside it, to produce Volterra’s coinage. The castle of Montieri stands at the foot of a hill with many mines in it. On the northern slopes of the hill, there also stood the site which we excavated. In the documents, this site was linked to the Bishop of Volterra and in the 1133 document it was called La Canonica dedicated to San Niccolò. The complex is composed of a church with six radiating conchs (Fig. 1c), located inside an enclosure that contained a central, open space with two long buildings, and a smaller construction, along its sides. Outside the enclosure there is a group of other structures, sited in relation to a further open space. The stratigraphy found so far bears witness to a settlement sequence spanning a long period of time [15]: initial occupation of the terraced site probably in the period between the 9th and

10th centuries; in the first half of 11th century the layout of the buildings at La Canonica comprised a church with six apses, with a small adjacent space containing a privileged burial. The peculiar shape of the church is likely to reflect the significance that this building must have had for the people who commissioned it. Indeed, as yet no parallels have been found in Italy, except for buildings dating to Late Antiquity. The objective of these operations would seem to have been the creation of an important centre of worship that was immediately venerated by local residents. This led to the formation of an area of lay burials arranged around the church between the 11th and 13th centuries. Excavation of less than a third of these burial areas, inside and outside the enclosure, has already brought to light more than 300 individuals belonging to the lay society of Montieri. The Bishop of Volterra’s objective

446

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

Fig. 4. Photomicrographs of the mortar aggregates, (a) sandstone in sample MM40, (b) rhyolite in sample MM7, (c) quartzite in sample MM7 and (d) cocciopesto in sample MM7.

was to combine control and management of mining resources with religious and social control over the Montieri community. This certainly strengthened the status of the Bishop himself, within the various dialectics of economic control of the local area on the part of the newly-formed local seigneurships. In the 12th century two long buildings and a stone enclosure were built. During the 12th century, and especially in the first decades of the 13th century, variations were made to the existing system of structures, and it was only at the end of 14th century that a new, residential building was added to the original structures. This was the last major intervention before final abandonment of the site, which occurred just before the 15th century.

(Li detector) (Figs. 6 and 7; Tables 5–7). The chemical analysis of the major (CaO, MgO, MnO, Fe2O3, P2O5, K2O, SiO2, Na2O, TiO2, Al2O3) and trace elements (Ni, Cu, Zn, Co, Cr, Pb, Ba, Sr, Rb, Nb, Y, Zr, V) on mortar pressed powders were carried out by a Philips MagiX X-ray fluorescence spectrometer (XRF), operating on Rh Ka tubes (Figs. 8 and 9; Tables 8 and 9). At the last, some variables were processed by a statistical discriminant analysis, using the software IBM SPSS STATISTICS 21.00.

2. Sampling and analytical techniques

The majority of the samples show light grey colour, main size class of the sand [16] and poor cohesion [17] (Fig. 2b). Lime lumps, coal fragments, straw, metallurgic slags and cocciopesto (fragments of pottery) were observed. Primary and secondary porosity, musk and biological patinas are often visible on the surface [18]. The mortars of the foundation walls are composed by a very poorly cohesive greyish-brown clay binder.

A total of 54 mortar samples were taken from the buildings of the archaeological site at the area 1000, area 2000, area 3000, area 4000 and area 9000 (Fig. 2; Table 1). The particularly interesting sampling was the area 1000, including the church (area 1000 – 2nd building) and its annex (area 1000 – 1st building). Petrographic, mineralogical and chemical characterisation required a multi-analytical approach. After a preliminary macroscopic analysis, the thin sections were observed with a polarised light optical microscope Leica DMRX (Figs. 3 and 4; Tables 2 and 3). Mineralogical composition was determined by a Bragg–Brentano Philips X’ Pert PRO PW 3040 X-ray diffractometer (XRD), equipped with a X’ Celerator PW 3015 detector. The instrument works on Cu Ka radiation, at 40 kV and 40 mA, and the diffraction data were collected in the range 3–60° 2h, step size of 0.02°, step time 58.42 s. The diffractograms were interpreted by using X’Pert High Score software (Fig. 5; Table 4). Textural and micro-chemical analysis of the binder (size <1/16), lumps, some minerals and lithic fragments in the aggregate were performed through a Philips XL30 FEI Quanta 200 electron scanning microscope (SEM), equipped with an EDAX Si

3. Results and discussion 3.1. Macroscopic analysis

3.2. Microscopic description on the thin section The microscopic analysis of the thin section under the polarised light optical microscope allowed us to describe the main features of each sample (Tables 2 and 3). The comparison between the samples let us to distinguish the following groups. 3.2.1. 1st Group (MM2, MM3, MM6, MM9, MM11, MM12, MM13, MM17, MM48_I, MM49, MM53, MM54) The mortars come from the six apses of the church. The poorly sorted aggregate has a main size of 450 lm (medium sand) [16], low sphericity and high roundness. It is mainly composed of shale (Fig. 3a), phyllites, sandstones, siltstones, limestones, quartzites

Table 2 Petrographic features of the mortar aggregate on thin sections. CS: coarse sand, MS: medium sand, FS: fine sand, PS: poorly sorted, LS: low sphericity, WR: well-rounded, SR: sub-rounded, SA: sub-angular, A: angular; R: random, P: preferential, H: homogeneous, I: inhomogeneous. Bt: biotite, Cal: calcite, Ms: muscovite, Opm: opaque mineral, Or: orthoclase, Opx: orthopyroxene, Pl: plagioclase, Qtz: quartz (abbreviations according to Refs. [20,21]). Aggregate Size (lm) Min.

Mean

4800

5

MM2

7000

10

MM3

5000

10

MM4

2400

10

MM5

2500

10

MM6

8000

10

MM7

2880

10

MM8

640

10

MM9

3000

10

MM10

1920

10

MM11

4560

10

MM12

4000

10

MM13

5500

10

MM14

1300

5

MM15

5000

5

MM16a

2960

5

MM16b

3500

5

MM17a

3200

5

MM17b

5000

10

MM18

3120

5

MM19

4700

5

MM20

3400

5

MM21

7000

5

MM22

2800

5

MM23

6000

10

160 FS 450 MS 450 MS 250 MS 650 CS 450 MS 650 CS 250 MS 450 MS 250 MS 450 MS 450 MS 450 MS 160 FS 160 FS 160 FS 160 FS 160 FS 450 MS 160 FS 250 MS 250 MS 250 MS 250 MS 400 MS

Sphericity

Roundness

Orientation

Distribution

Mineralogical phases

Rock fragments

Artificial stone material

Other

PS

LS

A-WR

R

I

PS

LS

WR

R

H

LS

WR

R

H

–

Metallurgic slag, charcoal, limestone partially burnt Charcoal, limestone partially burnt Charcoal

PS

LS

A-WR

R

I

Cocciopesto

Charcoal

PS

LS

A-WR

R

H

–

Metallurgic slag, charcoal

PS

LS

WR

R

H

–

Charcoal

PS

LS

A-WR

R

H

Cocciopesto

Metallurgic slag, charcoal

PS

LS

WR

R

I

Shale, phyllite, sandstone, siltstone, limestone, volcanic rock, marble, quartzite Shale, phyllite, sandstone, siltstone, limestone, flint, quartzite Shale, phyllite, sandstone, siltstone, limestone, quartzite Sandstone, shale, phyllite, siltstone, limestone, quartzite Limestone, marble, shale, phyllite, sandstone, volcanic rock, siltstone, quartzite Shale, phyllite, sandstone, siltstone, limestone, quartzite Shale, phyllite, sandstone, limestone, marble, volcanic rock, quartzite, flint Shale, sandstone

Cocciopesto, reused mortar -

PS

Qtz, Cal, Opm, Ms, Bt, Or Qtz, Cal, Ms, Opm Qtz, Cal, Pl, Ms, Opm Qtz, Cal, Pl, Opm

–

–

PS

LS

WR

R

H

–

Charcoal

PS

LS

A-WR

R

H

Shale, phyllite, sandstone, siltstone, limestone, quartzite, marble Quartzite, sandstone, shale, limestone

–

Glass

PS

LS

WR

R

H

–

Charcoal

PS

LS

WR

R

H

–

PS

LS

WR

R

H

Charcoal, limestone partially burnt Charcoal

PS

LS

A-WR

R

I

PS

LS

A-WR

R

I

Shale, phyllite, sandstone, siltstone, limestone, quartzite Shale, phyllite, sandstone, siltstone, limestone, quartzite Shale, phyllite, sandstone, siltstone, limestone, flint, quartzite Shale, phyllite, limestone, sandstone, volcanic rock, quartzite Shale, sandstone, limestone, siltstone, phyllite

PS

LS

WR

R

I

PS

LS

A-WR

R

I

PS

LS

WR

R

I

PS

LS

WR

R

H

PS

LS

WR

R

I

PS

LS

A

R

H

PS

LS

A

R

H

PS

LS

A

R

H

PS

LS

A

R

H

PS

LS

WR

R

H

Cal, Qtz, Or, Ms, Bt, Opm Qtz, Cal, Ms, Opm Cal, Qtz, Or, Ms, Bt, Opm Qtz, Opm Qtz, Cal, Pl, Ms, Bt, Opm Qtz, Opm, Cal, Ms Qtz, Cal, Or, Pl, Ms, Opm Qtz, Cal, Opm Qtz, Cal, Pl, Ms, Opm Qtz, Cal, Opm, Ms, Bt Qtz, Cal, Opm, Ms, Pl, Opx Ms, Bt, Qtz Qtz, Cal, Opm, Ms, Or, Bt Ms, Bt, Qtz, Cal, Opm Qtz, Cal, Ms, Bt, Opm Ms, Bt, Qtz, Cal, Opm Qtz, Cal, Opm, Ms Qtz, Cal, Opm, Or, Pl, Bt Qtz, Cal, Opm, Or Qtz, Cal, Opm, Or, Pl, Ms, Bt Qtz, Cal, Ms, Bt, Opm

– Reused mortar Cocciopesto

Metallurgic slag, charcoal, limestone partially burnt Metallurgic slag, charcoal

Sandstone, phyllite, shale, quartzite

–

–

Shale, phyllite, limestone, sandstone, siltstone, quartzite Phyllite, siltstone, sandstone, shale, limestone

Cocciopesto

Metallurgic slag, charcoal

–

Charcoal

Shale, phyllite, sandstone, siltstone, limestone, quartzite Phyllite, shale, sandstone, limestone, quartzite

–

Charcoal

–

Charcoal

Sandstone, quartzite, shale, limestone, siltstone, phyllite Sandstone, quartzite, limestone, shale, phyllite, siltstone Sandstone, quartzite, limestone, shale, phyllite, siltstone Sandstone, quartzite, limestone, shale, phyllite, siltstone Sandstone, shale, phyllite, siltstone, limestone

Cocciopesto

Charcoal

Cocciopesto

Charcoal

Cocciopesto

Charcoal

Cocciopesto

Charcoal

Reused mortar

Charcoal, limestone partially burnt, metallurgic salg 447

(continued on next page)

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

Max. MM1

Sorting

448

Table 2 (continued) Aggregate Size (lm) Min.

Mean

MM24

5000

5

MM25

3200

5

MM26a

3000

5

MM26b

2400

5

MM27

2000

5

MM28

2200

5

MM29a

2300

5

MM29b

4960

5

MM30

2050

5

MM31

4000

10

MM32

4000

10

MM33

8000

5

MM34

5000

10

MM35

4000

5

MM36

7000

5

MM37

8000

10

MM38

11000

10

MM39

5000

10

MM40

2800

10

MM41

3280

5

MM42_I

1840

10

MM42_II

1040

5

MM43

8000

5

MM44a

5000

10

MM44b_I

2680

5

MM44b_II

2480

5

250 MS 250 MS 250 MS 250 MS 250 MS 250 MS 160 FS 160 FS 160 FS 400 MS 400 MS 250 MS 400 MS 250 MS 250 MS 400 MS 400 MS 650 CS 651 CS 160 FS 400 MS 160 FS 160 FS 450 MS 160 FS 160 FS

Sorting

Sphericity

Roundness

Orientation

Distribution

Mineralogical phases

Rock fragments

Artificial stone material

Other

PS

LS

A

R

H

Charcoal

LS

A

R

H

–

–

PS

LS

A

R

H

Sandstone, quartzite, shale, phyllite, limestone, siltstone, marble Sandstone, quartzite, phyllite, shale, siltstone, limestone Quartzite, shale, phyllite

Cocciopesto

PS

–

–

PS

LS

A

R

H

Charcoal

LS

A

R

H

Sandstone, quartzite, phyllite, shale, siltstone, limestone Sandstone, quartzite, shale, limestone

–

PS

Cocciopesto

–

PS

LS

A

R

H

Sandstone, quartzite, phyllite, shale, limestone

Cocciopesto

Charcoal

PS

LS

A-WR

R

I

Shale, phyllite, limestone, sandstone, marble

–

PS

LS

WR

R

I

Qtz, Opm, Cal, Or, Pl, Ms, Bt Qtz, Opm, Cal, Or, Pl, Ms Qtz, Opm, Cal, Ms, Bt, Or Qtz, Opm, Cal, Or, Pl, Ms, Bt Qtz, Opm, Cal, Or, Pl, Ms, Bt Qtz, Cal, Opm, Or, Pl, Ms, Bt Cal, Opm, Qtz, Ms, Or, Bt Ms, Bt, Qtz

Phyllite, shale, sandstone

–

Metallurgic slag, charcoal, glass –

PS

LS

A-WR

R

I

Metallurgic slag, charcoal

LS

WR

R

H

Shale, phyllite, sandstone, siltstone, limestone, quartzite Shale, phyllite, sandstone, siltstone, limestone

Cocciopesto

PS

Reused mortar

PS

LS

WR

R

H

Shale, phyllite, sandstone, limestone

Reused mortar

PS

LS

A

R

H

Quartzite, sandstone, shale, phyllite, siltstone

Cocciopesto

Charcoal, limestone partially burnt Charcoal, limestone partially burnt Charcoal

PS

LS

WR

R

H

Shale, phyllite, sandstone, limestone

Reused mortar

PS

LS

A

R

I

–

LS

A

R

I

Sandstone, quartzite, shale, phyllite, siltstone, limestone Sandstone, quartzite, shale, phyllite, siltstone

Charcoal, limestone partially burnt Charcoal

PS

Cocciopesto

Charcoal

PS

LS

WR

R

H

Shale, phyllite, sandstone, limestone, siltstone

Reused mortar

PS

LS

WR

R

H

Shale, phyllite, limestone, sandstone, siltstone

Reused mortar

PS

LS

A-WR

R

H

Cocciopesto

LS

A-WR

R

H

Cocciopesto

Metallurgic slag, charcoal

PS

LS

WR

R

I

Shale, phyllite, limestone, sandstone, siltstone, volcanic rock Shale, phyllite, limestone, siltstone, marble, flint Sandstone, shale, phyllite, quartzite

Charcoal, limestone partially burnt Charcoal, limestone partially burnt Metallurgic slag, charcoal

PS

Cal, Opm, Qtz, Ms, Bt Qtz, Cal, Ms, Bt, Opm Qtz, Cal, Ms, Bt, Opm Qtz, Opm, Cal, Or, Ms Qtz, Cal, Ms, Bt, Opm Qtz, Opm, Cal, Or Qtz, Opm, Cal, Or, Pl, Ms, Bt Qtz, Cal, Ms, Opm Qtz, Cal, Ms, Opm Qtz, Cal, Opm, Or, Bt Qtz, Cal, Opm, Or, Ms, Bt Ms, Bt, Qtz, Or

–

Charcoal

PS

LS

WR

R

I

Charcoal

LS

A-WR

R

H

Sandstone, shale, phyllite, siltstone, limestone, quartzite Shale, phyllite, sandstone, siltstone

–

PS

Qtz, Opm, Bt, Cal Cal, Opm, Qtz

–

Charcoal

PS

LS

A-WR

R

I

Sandstone, phyllite, shale, limestone, siltstone

Reused mortar

PS

LS

WR

R

I

Sandstone, shale, phyllite, siltstone, limestone

–

Metallurgic slag, charcoal, limestone partially burnt Charcoal

PS

LS

A-WR

R

I

Metallurgic slag, charcoal

LS

WR

R

I

Phyllite, shale, sandstone, siltstone, marble, limestone Phyllite, shale, sandstone, limestone

–

PS

Qtz, Cal, Opm, Ms, Mc Ms, Bt, Qtz, Cal, Opm, Or Cal, Qtz, Opm, Ms, Bt Qtz, Cal, Ms, Opm

–

–

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

Max.

Cocciopesto

–

–

–

–

Shale, phyllite, volcanic rock, limestone

Shale, phyllite, sandstone, siltstone, limestone

Phyllite, shale, sandstone

Shale, phyllite, sandstone, siltstone, limestone

Shale, phyllite, sandstone, siltstone, limestone

Charcoal, limestone partially burnt

– Shale, phyllite, sandstone, siltstone, limestone

H WR PS

LS

R

H WR PS

LS

R

I WR PS

LS

R

H WR PS

LS

R

I R A-WR LS PS

WR PS

LS

R

H

Qtz, Cal, Ms, Opm Cal, Qtz, Opm, Ms Qtz, Cal, Ms, Opm Ms, Bt, Qtz, Cal, Pl Qtz, Pl, Ms, Bt, Opm Qtz, Ms, Bt, Opm

Charcoal

– Shale, phyllite, sandstone, siltstone, limestone H WR PS

LS

P

Charcoal, limestone partially burnt Charcoal

Metallurgic slag, charcoal, limestone partially burnt Charcoal, limestone partially burnt Charcoal – I A-WR PS

LS

R

Metallurgic slag, charcoal

Metallurgic slag, charcoal – A-WR

2240 MM54

10

9000 MM53

10

6000 MM52

5

10000 MM51

10

5000 MM50

10

2200 MM49

10

1600 MM48_I

10

1920 MM46

5

4000 MM45

5

160 FS 160 FS 450 MS 450 MS 650 CS 600 CS 160 FS 450 MS 451 MS

PS

LS

R

I

Cal, Qtz, Opm, Ms, Or, Bt Cal, Qtz, Opm, Ms, Or Qtz, Cal, Opm

Sandstone, phyllite, shale, limestone, siltstone, quartzite Phyllite, shale, limestone, sandstone, quartzite

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

449

and flints. A few monocrystalline phases, such as quartz, calcite, muscovite, opaque minerals, plagioclase, orthoclase and biotite, coal fragments (Fig. 3b) and traces of limestone partially burnt were identified. The aggregate amount, as assessed by visual estimation, is about 30% [19]. The binder shows a homogeneous greyish-brown colour and cryptocrystalline aspect; some lime lumps can be observed. The porosity is prevalently of secondary type and the main size of the pores is 200–250 lm. 3.2.2. 2nd Group (MM1, MM14, MM15, MM16b, MM29a, MM30, MM42_II, MM43, MM44b_I, MM45, MM46) The mortars mainly belong to the annex of the church (area 1000 – 1st building). Many monocrystalline phases such as calcite, quartz, euhedral opaque minerals, muscovite and lesser amounts of orthoclase, biotite, plagioclase, microcline and orthopyroxene were identified. Shales, sandstones, limestones, phyllites, siltstones, rhyolites, biomicrites, quartzites, metallurgical slags (Fig. 3c), coal, limestones partially burned, some reused mortars and rare glasses (Fig. 3d) were also recognised. The main size is about 160 lm (fine sand) [16] and the quantity of the aggregate is 40%. The porosity has a main size of about 120–200 lm and it was evaluated at 15%. 3.2.3. 3rd Group (MM19, MM20, MM21, MM22, MM24, MM25, MM26b, MM27, MM28, MM33, MM35, MM36) The samples were taken from the productive area (area 2000) and from the outer wall bounding the complex. The poorly sorted aggregate is mainly composed of quartz, sandstones and quartzites. The angular clasts show a main size of 250 lm (medium sand) [16]. Monocrystalline phases essentially are quartz, calcite, opaque minerals (Fig. 3e), iron oxides with growth structure (Fig. 3f), orthoclase, plagioclase, muscovite and biotite. In order of abundance, sandstones, quartzites, shales, limestones, siltstones, phyllites, fragments of pottery (cocciopesto) and coal can be observed. The aggregate was estimated at 50%. A further discrimination in the group arose between the mortar foundations (MM24, MM26b, MM27) and the samples of the outer wall (MM33, MM35, MM36). The first ones manifest an inhomogeneous binder and a higher aggregate amount, of about 50–55%; the second ones show a lower quantity of aggregate, of about 40%. 3.2.4. 4th Group (MM23, MM31, MM32, MM34, MM37, MM38) The mortars come from the square building (area 4000). A great number of partially burnt limestone and reused mortars were detected. The composition of the reused mortars was similar to that of the adjacent productive area (3rd group). Other lithic fragments are shales, phyllites, sandstones, limestones, siltstones, quartzites, few metallurgic slags and coal. The aggregate amount was evaluated at 35% and the porosity, of primary and secondary type, is about 15%. 3.2.5. 5th Group (MM5, MM7, MM39, MM40, MM50) The samples were taken from the north and south-west entrances of the church and from the area 9000. The aggregate displays a mean size of 650 lm (coarse sand) [16], medium–low sphericity and it is visually estimated at 35%. Several metallurgical slags are visible. Although the slags mainly have an opaque appearance in transmitted light, it was possible to recognise melilite crystals (Fig. 3g) and iron oxides, sometimes showing dendritic shape. Monocrystalline phases such as calcite (Fig. 3h), quartz, euhedral opaque minerals, muscovite, biotite and orthoclase compose the aggregate. Lithic fragments mainly consist of shales, sandstones (Fig. 4a), limestones, siltstones, phyllites, rhyolites (Fig. 4b), marble and quartzite (Fig. 4c); cocciopesto (Fig. 4d) and coal are also present. The primary and secondary porosity was valued at 15%.

450

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

Table 3 Main characteristics of the binder and porosity on thin sections and percentage amount of aggregate, binder and porosity by visual estimation. H: homogeneous, I: inhomogeneous, n.d.: not determined. Binder Colour

MM1

Porosity Crystallinity

Size (lm) Max.

Min.

Cryptocrystalline

4000

10

MM2

Pale Browngrayish Gray-brownish

Cryptocrystalline

3200

10

MM3

Gray-brownish

Cryptocrystalline

1680

10

MM4

Gray-brownish

Cryptocrystalline

4800

10

MM5

Brown-grayish

Cryptocrystalline

8000

10

MM6 MM7

Gray-brownish Brown-grayish

Cryptocrystalline Cryptocrystalline

4800 7000

10 10

MM8

Dark gray

Cryptocrystalline

1760

10

MM9 MM10 MM11 MM12 MM13 MM14 MM15 MM16a MM16b MM17a MM17b MM18 MM19

Gray-brownish Brown Gray-brownish Gray-brownish Gray-brownish Brown Brown-grayish Red-Brownish Brown Red-Brownish Gray-brownish Red-Brownish Pale brown

Cryptocrystalline Cryptocrystalline Cryptocrystalline Cryptocrystalline Cryptocrystalline Cryptocrystalline Cryptocrystalline – Cryptocrystalline – Cryptocrystalline – Cryptocrystalline

3200 3200 4400 2900 2600 1700 2500 1040 1800 1360 1500 1120 6900

10 10 10 10 10 10 10 10 10 10 10 10 10

MM20

Pale brown

Cryptocrystalline

1200

10

MM21

Brown-grayish

Cryptocrystalline

1080

10

MM22

Pale brown

Cryptocrystalline

1200

10

MM23

Brown

Cryptocrystalline

1600

10

MM24 MM25

Pale brown Gray-brownish

Cryptocrystalline Cryptocrystalline

1600 2800

10 10

MM26a MM26b MM27 MM28

Brown Pale brown Pale brown Pale brown

Cryptocrystalline Cryptocrystalline Cryptocrystalline Cryptocrystalline

1600 1680 3100 2200

10 10 10 10

MM29a MM29b MM30 MM31 MM32 MM33

Brown-grayish Red-Brownish Brown Brown Brown Pale brown

Cryptocrystalline – Cryptocrystalline Cryptocrystalline Cryptocrystalline Cryptocrystalline

1300 1440 1000 1900 2200 3000

10 10 10 10 10 10

MM34 MM35

Brown Pale brown

Cryptocrystalline Cryptocrystalline

2400 3200

10 10

MM36 MM37

Gray-brownish Brown

Cryptocrystalline Cryptocrystalline

2000 4000

10 10

MM38 MM39

Brown Gray-brownish

Cryptocrystalline Cryptocrystalline

3200 1840

10 10

MM40

Gray-brownish

Cryptocrystalline

2600

10

MM41 MM42_I

Red-Brownish Gray-brownish

– Cryptocrystalline

1356 3040

10 10

MM42_II

Gray-brownish

Cryptocrystalline

640

10

MM43 MM44a

Gray-brownish Gray-brownish

Cryptocrystalline Variable from micritic to microcrystalline

2320 880

10 10

% by visual estimation [16] Tipology

Distribution

Aggregate

Binder (size < 1/ 16 mm)

Porosity

Primary and secondary Primary and secondary Primary and secondary Primary and secondary Primary and secondary Secondary Primary and secondary Primary and secondary Secondary Primary Secondary Secondary Secondary Secondary Secondary Secondary Secondary Secondary Secondary Secondary Primary and secondary Primary and secondary Primary and secondary Primary and secondary Primary and secondary Secondary Primary and secondary Secondary Secondary Secondary Primary and secondary Secondary Secondary Secondary Secondary Secondary Primary and secondary Secondary Primary and secondary Secondary Primary and secondary Secondary Primary and secondary Primary and secondary Secondary Primary and secondary Primary and secondary Secondary Secondary

I

40

45

15

I

30

50

20

I

30

55

15

I

25

55

20

I

35

50

15

I I

30 35

50 50

20 15

I

3

82

15

I H I I I I I I I I I I H

30 40 30 30 30 40 40 30 40 n.d. 30 35 50

50 53 50 50 50 40 40 50 45 n.d. 50 50 43

20 7 20 20 20 20 20 20 15 n.d. 20 15 7

H

45

48

7

H

50

40

10

H

50

40

10

I

35

50

15

I H

50 50

35 40

15 10

I I I I

n.d. 50 50 50

n.d. 30 35 40

n.d. 20 15 10

I I I I I I

40 30 40 35 35 40

35 50 40 50 50 40

25 20 20 15 15 20

I I

35 40

50 45

15 15

I I

40 35

40 50

20 15

I H

35 35

50 55

15 10

I

35

50

15

I I

30 30

50 45

20 25

I

40

45

15

I I

40 40

45 40

15 20

451

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460 Table 3 (continued) Binder Colour

Porosity Crystallinity

Size (lm)

% by visual estimation [16] Tipology

Distribution

Aggregate

Binder (size < 1/ 16 mm)

Porosity

Max.

Min.

1000 1560

10 10

Secondary Secondary

I I

40 40

40 40

20 20

4800

10

Primary and secondary Primary and secondary Primary and secondary Primary and secondary Primary and secondary Primary and secondary Secondary Secondary Secondary

I

40

50

10

I

40

45

15

H

30

55

15

H

35

50

15

I

35

45

20

I

50

35

15

I I I

30 30 31

50 50 51

20 20 20

MM44b_I MM44b_II

Gray Brown

MM45

Gray-brownish

Cryptocrystalline Variable from micritic to microcrystalline Cryptocrystalline

MM46

Gray-brownish

Cryptocrystalline

1120

10

MM48_I

Gray-brownish

Cryptocrystalline

800

10

MM49

Gray-brownish

Cryptocrystalline

2400

10

MM50

Gray-brownish

Cryptocrystalline

4000

10

MM51

Gray-brownish

Cryptocrystalline

2400

10

MM52 MM53 MM54

Red-Brownish Gray-brownish Gray-brownish

– Cryptocrystalline Cryptocrystalline

1440 2240 8000

10 10 10

Fig. 5. Diffractogram of the sample MM7, Ak: akermanite, Cal: calcite, Chl: chlorite, Mgt: magnetite, Ms: muscovite, Qtz: quartz, Wus: wustite.

The mortars sampled from the foundation wall of the whole complex (MM16a, MM17a, MM18, MM29b, MM41, MM44b_II, MM47, MM52) are characterised by a clay binder. Among these samples, only in MM17a and MM18 were some lumps of lime identified. Other samples (MM4, MM8, MM10 and MM42_I, MM44a, MM51) showing different characteristics were not included in the groupings.

hydrothermal deposits, in association with sulfides. In other samples, a small percentage of magnesite (MM21, MM45), dolomite (MM50), pyroxene (MM51), fluorite (MM61) and clay minerals, such as kaolinite (MM23) and vermiculite (MM62), was found. Moreover, a detailed study was performed by microcapillary XRD analysis on the opaque minerals in thin sections. The main opaque phases consist of magnetite and goethite.

3.3. XRD analysis

3.4. SEM–EDS analysis

The main mineralogical phases detected by XRPD analysis were calcite, quartz, muscovite, clinochlore and albite (Table 4). In the samples of the 2nd and 5th groups, containing metallurgic slags, melilites (hardystonite, akermanite and gehlenite) and iron oxides (hematite, wustite and magnetite) were determined (Fig. 5). Very low amounts of smithsonite (ZnCO3) and proustite (Ag3AsS3) were also revealed in the samples MM15, MM16, MM36 and MM40. Smithsonite is commonly present as a product of oxidation in zinc deposits. Proustite is a secondary product in

The chemical composition of the binder was determined by SEM–EDS microanalysis. For each sample three spots were collected on a homogeneous area of the lumps and binder. A total of 153 analyses were performed on the binder and 149 on the lime lumps. The data were elaborated in order to obtain the average hydraulicity index for every sample. Tables 5 and 6, obtained by statistical treatment of the data, show the maximum, minimum and average values of the chemical elements and of the hydraulicity indexes for each groups, with the relative error, standard

452

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

Table 4 Mineralogical composition of mortars by X-ray powder diffraction; the number of stars is proportional to the phase abundancy. Ab: albite, Cal: calcite, Chl: chlorite, Hem: hematite, Mgt: magnetite, Mel: melilite, Ms: muscovite, Qtz: quartz, Rt: rutile, Wus: wustite (abbreviations according to Refs. [20,21]).

MM1 MM2 MM3 MM4 MM5 MM6_I MM6_II MM7 MM8 MM9 MM10 MM11 MM12 MM13 MM14 MM15 MM16 MM17 MM18 MM19 MM20 MM21 MM22 MM23 MM24 MM25 MM26 MM27 MM28 MM29 MM30 MM31 MM32 MM33 MM34 MM35 MM36 MM37 MM38 MM39 MM40 MM41 MM42 MM43 MM44 MM45 MM46 MM47 MM48_I MM48_II MM49 MM50 MM51 MM52 MM53 MM54

Cal

Qtz

Chl

Ms

***

***

**

*

***

***

**

*

*

***

***

**

**

*

***

***

**

**

***

***

*

**

***

***

**

**

*

***

***

*

**

*

***

***

**

*

***

***

**

**

***

***

**

**

***

***

**

*

***

***

**

*

*

***

***

**

**

*

***

***

**

**

*

***

***

*

*

***

***

*

*

***

***

**

*

*

***

***

**

**

*

***

***

**

**

*

***

***

**

**

*

***

***

*

*

***

***

*

**

***

***

*

**

***

***

**

**

***

***

*

**

***

***

*

**

***

***

*

**

***

***

*

*

*

***

***

*

**

*

***

**

**

*

***

***

*

*

***

**

*

*

***

**

*

*

***

***

*

*

***

***

**

*

***

***

*

**

***

***

*

**

***

**

*

*

***

***

**

**

***

***

**

**

***

***

*

*

***

**

**

*

***

***

**

*

*

**

***

**

**

*

***

**

*

*

*

***

***

*

**

*

***

***

*

**

***

***

**

**

***

**

**

*

***

***

**

*

***

***

**

**

***

***

*

*

***

***

**

**

***

**

**

**

***

***

**

**

*

***

***

**

**

*

deviation and variance. In the mortar samples of the 1st, 2nd, 3rd and 4th groups the lumps are composed by aerial lime and the binder shows weak hydraulicity [22]. In the mortars of the 5th group the lumps have weak hydraulicity and the binder is moderately hydraulic. The metallurgic slags and cocciopesto in the aggregate probably induced the highest hydraulicity of these ones. The binder in the samples MM10 and MM59 is eminently hydraulic and this is due to the presence of artificial glasses in the first case and of the great number of slags in the second one. Another interesting observation concerns the presence of the lead peaks in the binder and lumps microanalysis spectrums of

Ab

Mel

Mgt

Wus

Hem

*

**

Rt

*

*

**

*

**

*

** * *

*

*

*

*

*

* **

*

*

*

* * **

the 2nd and 5th groups. This can indicate the partial migration of this element from the slags into the surrounding area (Fig. 6). The composition of some metallurgic slags was studied by SEM–EDS microanalysis, confirming the presence of melilite, previously identified by optical microscopy and XRPD (Fig. 7a). Dendritic phases such as wustite or spinels (Fig. 7b), feldspathoids (Fig. 7c) and metal-bearing grains (Fig. 7d and e) were also detected (Table 6). The grains are constituted by lead, alloy, such as Cu–Sb and As–Pb–Sb, and copper sulfides. The morphological and chemical microanalysis was also carried out on the opaque minerals, previously studied by XRD microcapillary, to verify their

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

453

Fig. 6. SEM–EDS microanalysis of binder and lumps, (a) Back Scattered Electrons (BSE) image of lump and the corresponding SEM–EDS spectrum in the sample MM5, (b) BSE image of an area of the sample MM5 containing a metallurgic slag (indicated by the white arrow) and (c) map of the lead distribution.

composition (Fig. 7f). Furthermore particular polycrystalline argentiferous pyrite was detected in MM24 and MM48_II samples (Fig. 7g and h).

3.5. XRF analysis The chemical composition of the major (CaO, MgO, MnO, Fe2O3, P2O5, K2O, SiO2, Na2O, TiO2, Al2O3) and the trace elements (Ni, Cu, Zn, Co, Cr, Pb, Ba, Sr, Rb, Nb, Y, Zr, V) was obtained by XRF analysis (Tables 8 and 9). Among the major elements, in the foundation samples, the values of Fe, Si, Al and Ti are higher than in the other samples, due to the presence of a clay binder (Fig. 8a). The mortars containing metallurgic slags have the highest amount of Fe, Mg and Na, among the major elements, and Cu, Zn and Pb as traces (Fig. 8b).

In our case, the petrographic analysis on the thin sections highlighted some important compositional differences between the mortars, allowing us to make a preliminary grouping. Subsequent analysis led us to investigate the mineralogical (XRD) and chemical (SEM–EDS, XRF) composition, deepening and increasing the available data. A statistical discriminant analysis was performed using as variables the petrographic groups and the chemical data of the major elements. The groups are clearly identifiable in the graphic of the main functions (Fig. 9). The overlap of the 1st and 4th groups is the consequence of the very similar composition of the sand in the mortars. Petrographic, mineralogical and chemical investigations allowed us to highlight the compositional differences among the samples, favouring the reading of the construction steps of the site and supporting the archaeological interpretation. The 5 groups identified correspond to many construction phases (Fig. 10), listed below.

4. Reconstruction of the building phases 4.1. 1st Phase (1st group) Recent archaeometric researches showed how the study of the mortar composition can be of fundamental importance for the analysis of the masonry structures in ancient monuments, supporting the archaeological investigation in the reconstruction of the building phases [23–25]. In particular, the analysis of the mortars provides much information about the production technology. The workers operating in different historical periods probably produced building materials with different technological characteristics. The differences could be due to the type of the sand, the choice of limestone for the production of the binder, the addition of other natural or artificial materials (for example materials with hydraulic properties) and the processing. Sometimes, even within the same construction phase, differences can arise because of different or variable proportions between the binder and the aggregate.

In the first half of the XI century the building of the church took place. The mortar reveals similar characteristics in all the six apses that compose the edifice. The presence of the same mortar was traced in the semicircular wall remains inside the annex building (area 1000 – 1st building), probably contemporary with the church. The traces of walls in the area 2000, currently under further archaeological analyses, are probably also ascribed to the 1st phase. 4.2. 2nd Phase (2nd group) The building annexed to the church was erected in a consecutive step. Traces of the 1st group mortars were found in the foundations of the edifice and traces of the 2nd group mortars are

454

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

Fig. 7. BSE images, (a) metallurgic slag composed by melilite (Mel) and wustite (Wus) in sample MM1, (b, c) zoned spinels (Spl), Ni–Cu–Zn alloy and leucite in the slag of sample MM23, (d) composite metal-bearing grain with pure antimony (Sb) and Sb–Cu metal alloy in the slag of the sample MM7, (e) composite metal-bearing grain with pure lead (Pb) and Sb–Cu metal alloy in slag of the sample MM15, (f) iron oxide with structure of growth in sample MM20 and (g, h) argentiferous pyrite (Py).

455

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460 Table 5 Statistical data of SEM–EDS microanalysis calculated on the binder for each mortar group. H.I.: hydraulicity index, b.l.d.: below detection limit. SiO2

Al2O3

Fe2O3

MnO

MgO

CaO

Na2O

K2O

P2 O5

Cl2O

PbO2

SO3

ZnO

H.I.

1st Group

Average Standard deviation Variance Minimum Maximun

6.70 2.42 5.87 2.04 11.14

2.48 0.57 0.33 1.27 3.47

0.83 0.44 0.19 b.d.l. 1.37

0.22 0.46 0.21 b.d.l. 1.24

1.14 0.60 0.36 0.02 2.18

86.75 3.86 14.93 80.36 93.93

0.83 0.23 0.05 0.38 1.03

0.42 0.20 0.04 0.17 0.76

0.65 0.50 0.25 b.d.l. 1.19

b.d.l. – – – –

b.d.l. – – – –

b.d.l. – – – –

b.d.l. – – – –

0.12 0.04 0.00 0.04 0.19

2nd Group

Average Standard deviation Variance Minimum Maximun

5.83 1.02 1.03 3.89 7.29

2.66 0.51 0.26 2.05 3.82

1.20 0.35 0.12 0.63 1.91

b.d.l. – – – –

1.61 0.31 0.10 1.34 2.48

82.06 1.73 3.00 77.43 83.76

1.42 0.55 0.30 0.70 2.30

0.45 0.18 0.03 b.d.l. 0.72

1.05 0.53 0.28 b.d.l. 1.69

0.01 0.04 0.00 b.d.l. 0.13

3.71 2.02 4.07 1.03 6.85

b.d.l. – – – –

b.d.l. – – – –

0.12 0.02 0.00 0.08 0.14

3rd Group

Average Standard deviation Variance Minimum Maximun

5.99 2.26 5.09 2.51 11.04

2.51 0.63 0.39 1.52 3.68

1.97 1.50 2.25 0.61 5.23

0.27 0.45 0.20 b.d.l. 1.23

2.03 1.34 1.78 0.79 6.00

84.21 5.06 25.56 73.88 90.48

1.06 0.43 0.18 0.46 2.00

0.36 0.26 0.07 b.d.l. 0.67

0.57 0.56 0.31 b.d.l. 1.53

0.02 0.06 0.00 b.d.l. 0.21

0.63 0.87 0.75 b.d.l. 2.28

0.27 0.41 0.17 b.d.l. 1.26

0.09 0.29 0.08 b.d.l. 1.00

0.12 0.05 0.00 0.05 0.22

4th Group

Average Standard deviation Variance Minimum Maximun

7.68 0.91 0.83 6.63 9.09

2.89 0.87 0.76 1.24 3.94

1.26 0.17 0.03 1.03 1.45

b.d.l. – – – –

1.18 0.42 0.18 0.25 1.43

85.86 2.38 5.67 83.12 89.91

0.77 0.36 0.13 b.d.l. 1.17

0.12 0.15 0.02 b.d.l. 0.37

0.16 0.37 0.13 b.d.l. 0.98

b.d.l. – – – –

b.d.l. – – – –

0.08 0.13 0.02 b.d.l. 0.34

b.d.l. – – – –

0.14 0.02 0.00 0.10 0.16

5th Group

Average Standard deviation Variance Minimum Maximun

10.59 5.90 34.84 6.85 22.34

4.45 1.03 1.06 2.94 5.94

1.21 0.28 0.08 0.97 1.73

b.d.l. – – – –

6.76 2.52 6.34 4.55 11.52

71.23 9.95 98.92 52.01 79.02

1.64 0.52 0.27 1.25 2.60

0.31 0.17 0.03 b.d.l. 0.48

0.91 0.58 0.34 b.d.l. 1.77

0.07 0.14 0.02 b.d.l. 0.36

2.82 1.21 1.46 1.72 4.75

b.d.l. – – – –

b.d.l. – – – –

0.22 0.12 0.02 0.13 0.46

Table 6 Statistical data of SEM–EDS microanalysis calculated on the lumps for each mortar group. H.I.: hydraulicity index, b.l.d.: below detection limit. SiO2

Al2O3

Fe2O3

MnO

MgO

CaO

Na2O

K2O

P2O5

Cl2O

PbO2

SO3

H.I.

1st Group

Average Standard deviation Variance Minimum Maximun

3.34 1.16 1.35 1.18 5.01

1.48 0.55 0.31 0.67 2.34

0.64 0.16 0.03 0.44 1.01

0.21 0.45 0.20 b.d.l. 1.24

1.03 0.45 0.20 0.18 1.51

91.40 2.38 5.64 87.61 94.99

0.87 0.18 0.03 0.59 1.14

0.29 0.14 0.02 0.08 0.54

0.75 0.57 0.33 b.d.l. 1.38

b.d.l. – – – –

b.d.l. – – – –

b.d.l. – – – –

0.06 0.02 0.00 0.03 0.09

2nd Group

Average Standard deviation Variance Minimum Maximun

2.93 1.44 2.08 1.18 5.19

1.49 0.65 0.43 0.72 2.58

0.79 0.29 0.08 0.47 1.36

b.d.l. – – – –

1.70 0.55 0.31 0.83 2.74

87.84 3.93 15.42 82.76 94.67

1.40 0.68 0.46 0.39 2.55

0.29 0.16 0.03 b.d.l. 0.47

0.94 0.62 0.38 b.d.l. 1.69

0.05 0.14 0.02 b.d.l. 0.47

2.58 1.40 1.97 0.76 5.52

b.d.l. – – – –

0.06 0.03 0.00 0.03 0.10

3rd Group

Average Standard deviation Variance Minimum Maximun

2.87 1.48 2.20 1.22 5.60

1.27 0.44 0.20 0.87 2.44

0.90 0.25 0.06 0.49 1.21

b.d.l. – – – –

1.72 0.68 0.47 0.47 3.20

90.50 3.05 9.28 85.41 95.15

1.10 0.44 0.19 0.39 1.91

0.12 0.16 0.03 b.d.l. 0.40

0.58 0.65 0.42 b.d.l. 1.55

0.13 0.24 0.06 b.d.l. 0.81

0.51 0.66 0.43 b.d.l. 1.88

0.31 0.56 0.32 b.d.l. 1.55

0.06 0.02 0.00 0.03 0.09

4th Group

Average Standard deviation Variance Minimum Maximun

3.11 0.62 0.39 1.91 3.91

1.26 0.27 0.07 0.85 1.61

0.82 0.13 0.02 0.59 1.01

b.d.l. – – – –

1.30 0.71 0.50 0.71 2.81

92.64 1.60 2.56 90.22 95.26

0.55 0.16 0.02 0.38 0.85

0.04 0.06 0.00 b.d.l. 0.14

0.19 0.43 0.18 b.d.l. 1.15

0.01 0.01 0.00 b.d.l. 0.04

b.d.l. – – – –

0.09 0.13 0.02 b.d.l. 0.30

0.06 0.01 0.00 0.04 0.07

5th Group

Average Standard deviation Variance Minimum Maximun

8.10 1.52 2.30 5.79 9.86

2.68 0.49 0.24 2.06 3.35

0.71 0.27 0.07 0.42 1.00

b.d.l. – – – –

5.24 0.98 0.96 3.73 6.30

78.36 3.52 12.36 73.43 83.37

1.22 0.40 0.16 0.62 1.72

0.25 0.05 0.00 0.18 0.31

0.76 0.50 0.25 b.d.l. 1.36

0.04 0.07 0.01 b.d.l. 0.17

2.65 0.60 0.36 2.15 3.65

b.d.l. – – – –

0.14 0.02 0.00 0.11 0.17

present in the foundations of the south-east apse of the church. In this phase, the presence of slags in the aggregate is widespread.

mostly for what concerns the type of aggregate, mainly composed by quartz. This feature should indicate the change of the place for the supply of sand.

4.3. 3rd Phase (3rd group) 4.4. 4th Phase (4th group) The 3rd phase coincides with the expansion of the productive area and the construction of a boundary wall. The composition of the mortar looks very different compared to the previous walls,

In the XIII century a quadrangular building (area 4000) was erected. The posteriority of the building is suggested both by

456

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

Table 7 SEM–EDS analyses of the pyrometallurgic phases in the slags (MM_S: Montieri Mortar_Slag); the chemical formula was recalculated to 7 oxygens for melilite and to 6 oxygens for leucite. Melilites

SiO2 TiO2 MgO Al2O3 FeO MnO CaO Na2O K2O ZnO Si Ti Mg Al Fe Mn Ca Na K Zn

Leucite

MM1a_S1_1

MM7a_S1 _1

MM7b_S2_1

MM7b_S2_2

MM23_S1_3

MM23_S1_4

35.54 – 3.39 10.98 7.66 – 38.24 3.62 0.57 – 1.69 – 0.24 0.62 0.31 – 1.95 0.33 0.03 –

36.14 – 3.34 12.67 7.00 – 32.88 3.43 0.39 4.14 1.72 – 0.24 0.71 0.28 – 1.67 0.32 0.02 0.15

31.79 – 2.34 17.98 5.07 – 35.78 2.50 – 4.54 1.51 – 0.17 1.01 0.20 – 1.82 0.23 – 0.16

28.43 – 2.69 11.46 16.86 – 31.45 3.41 – 5.70 1.35 – 0.19 0.64 0.67 – 1.60 0.31 – 0.20

56.83 – – 22.79 1.35 – – 0.33 18.70 – 2.04 – – 0.96 0.04 – – 0.02 0.86 –

56.99 – – 22.12 1.87 – – 0.82 18.21 – 2.05 – – 0.94 0.06 – – 0.06 0.83 –

Table 8 Chemical composition of mortars by XRF analysis: the major elements are expressed in oxides wt%. L.O.I.: loss on ignition. wt%

SiO2

TiO2

Al2O3

Fe2O3

MnO

MgO

CaO

Na2O

K2O

P 2 O5

L.O.I.

MM1 MM2 MM3 MM4 MM5 MM6_I MM6_II MM7 MM8 MM9 MM10 MM11 MM12 MM13 MM14 MM15 MM16 MM17 MM18 MM19 MM20 MM21 MM22 MM23 MM24 MM25 MM26 MM27 MM28 MM29 MM30 MM31 MM32 MM33 MM34 MM35 MM36

32.51 33.39 36.19 19.61 27.29 41.30 32.27 32.26 15.37 34.08 43.83 35.11 31.27 33.68 22.13 28.69 30.62 49.04 55.10 42.77 45.88 54.49 44.59 28.83 48.12 44.93 47.25 57.51 58.23 56.05 27.31 13.94 22.37 37.36 34.45 51.19 51.37

0.34 0.33 0.34 0.21 0.30 0.42 0.30 0.38 0.17 0.37 0.23 0.35 0.30 0.32 0.28 0.38 0.33 0.65 0.76 0.19 0.20 0.19 0.20 0.31 0.20 0.17 0.29 0.28 0.26 0.86 0.31 0.13 0.25 0.21 0.26 0.22 0.21

8.41 9.67 9.81 5.91 7.27 11.74 7.32 8.40 4.97 10.55 6.41 9.94 9.40 9.65 7.19 8.21 7.98 15.21 17.04 5.77 6.08 5.83 5.88 8.94 6.10 5.53 8.26 7.91 7.72 18.28 7.86 4.32 6.48 6.41 7.30 6.43 6.39

5.60 4.15 4.33 2.96 6.58 4.66 8.48 9.06 2.32 4.27 3.04 4.40 3.89 4.00 5.02 8.12 8.32 7.10 7.45 2.81 3.41 2.92 3.64 3.87 3.21 4.60 4.09 4.25 4.19 8.31 4.42 1.65 4.46 3.53 3.37 3.82 3.74

0.19 0.10 0.08 0.12 0.26 0.09 0.14 0.42 0.09 0.09 0.26 0.09 0.09 0.09 0.19 0.30 0.27 0.21 0.24 0.19 0.19 0.22 0.23 0.13 0.22 0.20 0.33 0.30 0.16 0.22 0.15 0.06 0.14 0.15 0.11 0.21 0.19

2.94 1.67 1.93 1.66 7.82 1.77 2.23 6.37 2.72 1.81 2.68 1.72 1.82 1.86 2.20 2.93 2.89 2.22 2.26 1.68 1.48 1.14 1.52 2.07 1.38 1.36 1.49 2.09 1.30 2.26 2.42 1.40 2.33 2.03 2.03 1.28 1.56

25.04 24.70 22.79 41.95 25.98 19.58 24.17 22.98 37.15 23.36 22.44 23.23 25.88 24.58 32.85 26.20 25.24 9.34 4.19 24.86 22.61 18.33 24.04 28.11 21.82 23.39 18.78 13.29 13.49 2.33 28.92 42.44 33.77 26.40 26.03 19.12 18.31

0.58 0.30 0.32 0.13 0.39 0.42 0.21 0.53 0.09 0.30 0.32 0.30 0.26 0.29 0.25 0.66 0.74 0.45 0.50 0.12 0.15 0.23 0.22 0.22 0.19 0.17 0.21 0.23 0.24 0.38 0.35 0.05 0.16 0.14 0.15 0.16 0.17

1.05 1.16 1.18 0.59 0.84 1.50 1.02 0.91 0.46 1.26 0.95 1.13 1.10 1.15 0.89 1.16 1.12 2.44 3.08 0.98 1.03 1.04 1.07 1.03 1.12 0.96 1.44 1.47 1.53 3.32 1.01 0.31 0.79 0.89 0.87 1.14 1.23

0.14 0.07 0.08 0.08 0.13 0.10 0.08 0.16 0.08 0.07 0.09 0.08 0.06 0.06 0.15 0.15 0.37 0.28 0.19 0.15 0.10 0.10 0.12 0.10 0.16 0.15 0.48 0.30 0.16 0.23 0.13 0.06 0.12 0.12 0.14 0.24 0.18

23.21 24.47 22.95 26.77 23.14 18.42 23.80 18.52 36.58 23.86 19.75 23.66 25.93 24.32 28.85 23.21 22.11 13.05 9.19 20.47 18.86 15.50 18.50 26.38 17.49 18.55 17.38 12.36 12.72 7.78 27.12 35.65 29.12 22.76 25.30 16.20 16.66

457

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460 Table 8 (continued) wt%

SiO2

TiO2

Al2O3

Fe2O3

MnO

MgO

CaO

Na2O

K2O

P 2 O5

L.O.I.

MM37 MM38 MM39 MM40 MM41 MM42 MM43 MM44 MM45 MM46 MM47 MM48_I MM48_II MM49 MM50 MM51 MM52 MM53 MM54

26.20 25.41 34.57 36.21 58.15 32.35 32.16 34.97 28.37 24.65 47.88 21.27 21.21 34.78 37.48 46.75 59.09 47.01 34.87

0.25 0.26 0.46 0.45 0.88 0.37 0.41 0.41 0.34 0.32 0.61 0.20 0.12 0.35 0.40 0.60 0.94 0.44 0.42

6.99 7.52 10.17 9.91 18.00 9.65 9.99 10.17 8.23 8.10 14.65 6.14 3.91 10.31 8.75 14.88 17.77 11.99 11.51

3.15 3.76 8.61 9.70 7.77 4.59 5.70 9.58 6.44 4.01 7.67 2.32 1.67 4.22 8.80 7.01 7.54 4.79 5.07

0.12 0.13 0.36 0.47 0.26 0.14 0.19 0.30 0.25 0.15 0.35 0.08 0.07 0.11 0.37 0.12 0.32 0.15 0.15

2.05 2.18 3.90 5.01 2.21 2.40 3.42 2.41 2.46 2.90 1.89 2.76 4.62 1.86 8.33 2.22 2.01 1.63 1.89

31.76 30.98 21.90 19.68 0.65 24.40 23.68 21.16 26.64 30.05 10.19 34.25 35.66 23.16 16.83 12.21 1.00 15.29 22.07

0.17 0.17 0.50 0.63 0.41 0.40 0.52 0.40 0.68 0.18 0.42 0.12 0.07 0.28 0.71 0.59 0.36 0.41 0.31

0.74 0.82 1.27 1.26 3.35 1.35 1.41 1.38 1.04 1.21 2.24 0.67 0.41 1.22 1.04 2.18 3.34 1.70 1.46

0.11 0.10 0.33 0.26 0.26 0.18 0.30 0.20 0.18 0.12 0.12 0.10 0.06 0.06 0.17 0.11 0.31 0.09 0.07

28.45 28.67 17.93 16.42 8.04 24.18 22.22 19.02 25.36 28.31 13.98 32.10 32.21 23.65 17.12 13.33 7.32 16.50 22.17

Fig. 8. Elaboration of the chemical data, (a) Al2O3 vs Fe2O3 diagram, and (b) Pb vs Cu diagram.

4.5. 5th Phase (5th group) Finally, a subsequent reconstruction by bricks of the entrances can be observed. The mortars, light in colour, contain many fragments of metallurgic slags of high dimensions.

5. Conclusions

Fig. 9. Discriminant analysis, Function 1 vs Function 2 diagram, obtained by using the petrographic groups and chemical data as variables.

stratigraphic relationships and the presence of reused mortars similar to those of the adjacent production area (3rd group). The limestones partially burnt in the mortar aggregates reveal a regress of the technical execution.

Nowadays the archaeological research more and more needs the contribution of archaeometry, most of all when the study objects are the construction materials (natural stones, mortars, plasters, bricks, etc.). The study of this kind of materials can provide several useful information for reconstructing the history of an archaeological site and in this context the multi-disciplinary and multi-analytical approach is fundamental. The study we performed allowed us to determine the petrographic, mineralogical and chemical composition of the mortar samples coming from the St. Niccolò ecclesiastical complex in Montieri. In particular, the petrographic description revealed the analogies and differences among the samples, enabling the identification of 5 typologies of mortars. The mineralogical and chemical data confirmed the differences, clearly identifiable in the graph of statistical discriminant analysis (Fig. 9). The integration of the archaeological and

458

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

Table 9 Trace element compositions of mortars by XRF analysis, expressed in ppm. L.O.I.: loss on ignition, b.d.l.: below detection limit.

MM1 MM2 MM3 MM4 MM5 MM6_I MM6_II MM7 MM8 MM9 MM10 MM11 MM12 MM13 MM14 MM15 MM16 MM17 MM18 MM19 MM20 MM21 MM22 MM23 MM24 MM25 MM26 MM27 MM28 MM29 MM30 MM31 MM32 MM33 MM34 MM35 MM36 MM37 MM38 MM39 MM40 MM41 MM42 MM43 MM44 MM45 MM46 MM47 MM48_I MM48_II MM49 MM50 MM51 MM52 MM53 MM54

Ni

Cu

Zn

Co

Cr

Pb

Ba

Sr

Rb

Nb

Y

Zr

V

46 31 33 26 46 36 31 55 20 33 31 32 33 33 34 72 44 65 76 23 20 20 22 36 20 20 36 32 21 91 34 15 27 28 30 24 20 24 34 70 73 86 39 63 58 49 30 62 20 15 37 64 61 76 40 44

2503 46 40 285 2868 101 1050 4779 429 56 1253 48 65 40 1646 2660 3375 113 109 608 740 746 773 125 1257 1552 484 963 844 300 1590 59 298 413 221 801 1187 123 135 4576 5383 68 887 1961 6419 3988 329 68 134 169 40 4200 346 50 109 59

10,539 71 66 1684 8822 179 2291 12,893 554 81 6483 70 66 60 3608 15,736 14,348 343 302 1574 2000 5609 2814 314 1872 2800 1451 2758 4341 1724 4281 97 1318 1170 528 2130 2396 384 324 10,154 13,584 215 3965 4753 7006 16,259 2252 149 300 464 67 15,204 1139 160 1188 100

11 12 8 6 8 11 11 12 4 11 9 12 10 11 8 14 9 21 26 10 9 12 10 8 10 12 17 14 10 30 6 2 6 10 8 11 9 9 8 13 12 25 11 13 12 9 9 22 4 4 10 11 15 31 15 13

62 48 53 24 60 62 51 77 26 57 37 52 47 47 41 62 57 119 151 29 27 31 31 48 29 25 50 42 41 183 48 22 40 34 47 34 35 38 43 91 91 182 59 83 82 54 44 102 32 18 55 80 94 176 68 66

17136 60 34 1720 12094 2878 4707 12186 2461 56 7780 49 65 36 14768 16893 29061 249 132 2859 3486 3408 4283 837 6945 4971 2605 3960 3447 1131 20029 258 2352 2564 1076 3512 4140 802 562 14749 12511 118 12234 15649 21182 18544 1480 74 1241 1197 36 12198 3496 47 86 95

b.d.l. 227 229 86 b.d.l. 258 b.d.l. b.d.l. 137 197 b.d.l. 226 240 209 30 b.d.l. b.d.l. 391 463 92 59 b.d.l. 4 226 2 5 124 62 b.d.l. 466 51 155 158 156 173 67 47 176 178 102 57 444 160 141 89 b.d.l. 157 409 164 94 227 8 381 526 223 251

107 145 98 118 141 144 124 210 96 132 46 119 118 132 110 92 85 124 89 108 101 65 79 290 90 80 427 86 55 55 142 292 384 244 384 81 105 327 486 113 263 65 138 76 132 200 92 94 251 213 106 171 160 56 97 142

3782 65 61 405 2667 690 1050 2688 566 68 1725 65 65 60 3263 3729 6385 167 170 662 795 777 969 236 1543 1110 625 908 796 392 4415 80 555 600 281 803 942 217 170 3264 2764 186 2729 3473 4669 4087 376 117 307 284 63 2691 855 183 93 92

9 11 12 9 8 11 24 10 8 11 8 11 11 11 9 10 8 15 16 7 6 6 7 11 6 6 9 7 7 17 9 5 10 7 9 6 6 10 10 12 10 16 11 11 9 9 12 14 9 5 10 9 14 18 12 13

9 8 7 7 11 8 5 12 5 7 12 7 7 8 8 9 9 11 12 8 9 8 11 10 12 10 13 11 10 10 8 7 10 9 11 9 9 11 12 12 15 14 7 10 10 10 8 11 6 6 7 13 13 10 9 9

64 71 64 36 54 89 31 65 34 59 40 60 56 57 52 64 61 114 131 36 35 29 34 68 37 26 79 44 38 151 62 39 65 48 72 35 39 62 81 69 76 160 71 78 70 66 62 96 48 32 56 64 101 101 101 101

53 58 59 28 52 73 59 68 26 65 55 61 54 54 43 60 51 123 151 46 43 41 45 51 40 35 63 57 62 179 49 20 39 36 47 46 51 37 42 83 103 183 54 67 72 57 42 119 31 23 60 72 108 181 88 79

archaeometric data leads to the recognition of the 5 major building phases (Fig. 10). The 1st phase coincides with the construction of the church, in the first half of the XI century. Between the late XI and early XII century other edifices were erected. The building technique seems to change over time: in the annex to the church it was possible to identify the presence of metallurgical slags and cocciopesto in the aggregate of the mortars. The slags are indicative of metallurgical processes that took place in this period in an area characterised by abundant mineral resources. Both slags and cocciopesto give hydraulic properties to the binder. In addition to the use of artificial materials in the aggregate, the variability of the aggregate-binder proportion and of the aggregate typologies, particularly evident in the area 2000, testify to the changes of the production technologies. The square building (area 4000) was

probably built in the XIII century. In this building, the mortars reveal the worst executive technique, due to the presence of numerous limestones partially burnt, reused mortars and the highest porosity. The analysis of some metallurgical slags, consisting of polymetallic grains of Cu, Pb, Sb and As, suggest the exploitation of mixed sulphide mineralisations [26,27]. Furthermore, the identification of argentiferous pyrite testifies to the mineral wealth of the place, in which the historical documents attest to the extraction of silver for the production of coins. The characterisation of mortar samples is the starting point for the study of provenance of the raw materials, the object of a future work. Furthermore, the knowledge of the original materials can be useful for the reproduction of mortars highly compatible, for the

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

459

Fig. 10. Location of the samples belonging to different phases of construction: a symbol corresponds to each phase.

proper restoration and the long-term conservation of the archaeological site [28,29], using mixing optimisation methods on compositional chemical data [30–33]. Acknowledgements The authors would like to thank the Superintendence for Archaeological Heritage of Tuscany and the Medieval Archaeology group of Siena University, for allowing us the sampling of mortars from the archaeological site, and Andrea Scala and the laboratory technicians of Physical, Earth and Environmental Sciences Department, for their contribution during sampling and analysis. The authors also thank the Science for Conservation of Cultural

Heritage PhD course of Florence University for funding the research. References [1] UNI, Cultural heritage. Mortars for building and decorative elements. Classification and terminology (N.10924), UNI, Milan, 2001. [2] A. Bakolas, G. Biscontin, A. Moropoulou, E. Zendri, Characterization of the lumps in the mortars of historic masonry, Thermochim. Acta 269–70 (1995) 809–816. [3] A. Moropoulou, A. Bakolas, K. Bisbikou, Investigation of the technology of historic mortars, J. Cult. Herit. 1 (2000) 45–58. [4] A. Moropoulou, K. Polikreti, A. Bakolas, P. Michailidis, Correlation of physicochemical and mechanical properties of historical mortars and classification by multivariate statistics, Cem. Concr. Res. 33 (2003) 891–898.

460

N. Chiarelli et al. / Construction and Building Materials 96 (2015) 442–460

[5] G.M. Crisci, M. Franzini, M. Lezzerini, T. Mannoni, M.P. Riccardi, Ancient mortars and their binder, Period. Mineral. 73 (2004) 259–268. [6] J. Elsen, Microscopy of historic mortars – a review, Cem. Concr. Res. 36 (2006) 1416–1424. [7] F. Carò, M.P. Riccardi, M.T. Mazzilli Savini, Characterization of plasters and mortars as a tool in archaeological studies: the case of Lardirago castle in Pavia, Northern Italy, Archaeometry 50 (2008) 85–100. [8] S. Pavìa, S. Caro, An investigation of Roman mortar technology through the petrographic analysis of archaeological material, Constr. Build. Mater. 22 (2008) 1807–1811. [9] D. Miriello, A. Bloise, G.M. Crisci, E. Barrese, C. Apollaro, Effect of milling: a possible factor influencing the durability of historical mortars, Archaeometry 52 (2010) 668–679. [10] M.P. Riccardi, P. Duminuco, C. Tomasic, P. Ferlonic, Thermal, microscopic and X-ray diffraction studies on some ancient mortars, Thermochim. Acta 321 (1998) 207–214. [11] J. Sanjurjo-Sánchez, M.J. Trindade, R. Blanco-Rotea, R. Benavides Garcia, D. Fernández Mosquera, et al., Chemical and mineralogical characterization of historic mortars from the Santa Eulalia de Bóveda temple, NW Spain, J. Archaeol. Sci. 37 (2010) 2346–2351. [12] A. Santos Silva, T. Cruz, M.J. Paiva, A. Candeias, P. Adriano, et al., Mineralogical and chemical characterization of historical mortars from military fortifications in Lisbon harbour (Portugal), Environ. Earth Sci. 63 (2011) 1641–1650. [13] G. Bianchi, Curtes, castelli e comunità rurali di un territorio minerario toscano. Nuove domande per consolidati modelli, in: P. Galetti (Ed.), Proceedings of the International Congress on Villaggi, comunità, paesaggi medievali; 2010 January 14–16, Fondazione Centro Italiano di Studi sull’alto Medioevo, Bologna, Italy. Spoleto, 2012, pp. 495–510. [14] G. Bianchi, J. Bruttini, L. Dallai, F. Grassi, Nuovi dati dalla ricerca archeologica per la ricostruzione del paesaggio storico delle Colline Metallifere massetane, in: G. Galeotti, M. Paperini (Eds.), Città e territorio. Conoscenza, tutela e valorizzazione dei paesaggi culturali, Debatte, Livorno, 2013, pp. 81–85. [15] G. Bianchi, J. Bruttini, F. Grassi, Lo scavo della Canonica di san Niccolò a Montieri (Gr), in: Notiziario della Soprintendenza per i Beni Archeologici della Toscana, 8/2012, All’Insegna del Giglio, Firenze, 2013, pp. 564–567. [16] C.K. Wentworth, A scale of grade and class terms for clastic sediments, J. Geol. 30 (1922) 377–392. [17] UNI, Cultural heritage. Petrographic description of a mortar (N. 11176), UNI, Milan, 2006. [18] UNI, Cultural heritage. Natural and artificial stone. Description of the alteration – terminology and definition (N. 11182), UNI, Milan, 2006. [19] L. Baccelle, A. Bosellini, Diagrammi per la stima visiva della composizione percentuale nelle rocce sedimentarie, in: Annali Geologiche e Paleontologiche Scienze, Sez. IX, Ferrara Univ, 1965, pp. 59–62.

[20] R. Kretz, Symbols for rock-forming minerals, Am. Mineral. 68 (1983) 277–279. [21] F.S. Spea, Metamorphic Phase-equilibria and Pressure–Temperature–Time Paths, Mineralogical Society of America, Washington DC, 1993. [22] E. Pecchioni, F. Fratini, E. Cantisani, Le malte antiche e moderne tra tradizione ed innovazione, fifth ed., Patron Ed, Bologna, 2013. [23] D. Miriello, D. Barca, A. Bloise, A. Ciarallo, G.M. Crisci, et al., Characterization of archaeological mortars from Pompeii (Campania, Italy) and identification of construction phases by compositional data analysis, J. Archaeol. Sci. 37 (2010) 2207–2223. [24] 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) 794–804. [25] R. De Luca, M.A. Cau Ontiveros, D. Miriello, A. Pecci, E. Le Pera, et al., Archaeometric study of mortars and plasters from the Roman City of Pollentia (Mallorca-Balearic Islands), Period. Mineral. 82–3 (2013) 353–379. [26] A. Manasse, M. Mellini, Chemical and textural characterisation of medieval slags from the Massa Marittima smelting sites (Tuscany, Italy), J. Cult. Herit. 3 (2002) 187–198. [27] M. Benvenuti, G. Bianchi, J. Bruttini, M. Buonincontri, L. Chiarantini, et al., Studying the Colline Metallifere mining area in Tuscany: an interdisciplinary approach, in: Joseph Silvertant (Ed.), Proceedings of the 9th International Symposium on Archaeological Mining History; 2014 June 5–8; Trento, Italy, Silvertant, Valkenburg, 2014, pp. 261–287. [28] L. Schueremans, Ö. Cizero, E. Janssens, G. Serré, K. Van Balen, Characterization of repair mortars for the assessment of their compatibility in restoration projects: research and practice, Constr. Build. Mater. 25 (2011) 4338– 4350. [29] TC 203-RHM, Repair mortars for historic masonry. Testing of hardened mortars, a process of questioning and interpreting, Mater. Struct. 42 (2009) 853–865. [30] D. Miriello, G.M. Crisci, Mixing and provenance of raw materials in the bricks from the Svevian castle of Rocca Imperiale (North Calabria – Italy), Eur. J. Mineral. 19 (2007) 137–144. [31] D. Miriello, D. Barca, G.M. Crisci, L. Barba, J. Blancas, et al., Characterization and provenance of lime plasters from the Templo Mayor of Tenochtitlan (Mexico City), Archaeometry 53 (2011) 1119–1141. [32] D. Miriello, M. Lezzerini, F. Chiaravalloti, A. Bloise, C. Apollaro, et al., Replicating the chemical composition of the binder for restoration of historic mortars as an optimization problem, Comput. Concr. 12 (2013) 553– 563. [33] D. Miriello, A. Bloise, R. De Luca, C. Apollaro, G.M. Crisci, et al., First compositional evidences on the local production of Dressel 2–4 amphorae in Calabria (Southern Italy): characterization and mixing simulations, Appl. Phys. A: Mater. Sci. Process. 119 (2015) 1595–1608.