The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy)

G Model CULHER-3605; No. of Pages 12

ARTICLE IN PRESS Journal of Cultural Heritage xxx (2019) xxx–xxx

Available online at

ScienceDirect www.sciencedirect.com

Original article

The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy)夽 Concetta Rispoli a,∗ , Alberto De Bonis a , Vincenza Guarino a , Sossio Fabio Graziano a,b , Claudia Di Benedetto a , Renata Esposito c , Vincenzo Morra a , Piergiulio Cappelletti a a Dipartimento di Scienze della Terra, dell’Ambiente e delle Risorse, Università di Napoli Federico II, Complesso Universitario Monte Sant’Angelo, Ed. L, Via Cintia 26, 80126 Naples, Italy b Dipartimento di Farmacia, Università degli Studi di Napoli Federico II, Via D. Montesano 49, 80131 Napoli, Italy c Dipartimento di Studi Umanistici, Università degli Studi di Napoli Federico II, Via Porta di Massa 1, 80133 Napoli, Italy

a r t i c l e

i n f o

Article history: Received 18 January 2019 Accepted 17 May 2019 Available online xxx Keywords: Mortar Roman Hydraulicity Index C-A-S-H gel Terme di Baia Campi Flegrei

a b s t r a c t Ancient pozzolanic mortars show the high technological quality achieved by Roman construction workers, due to their ‘excellent state’ of preservation in every environment. These workers well knew that thanks to the combination of lime with specific volcanic products (pozzolana), mortar and concrete become hydraulic, allowing underwater hardening and increasing mechanical strength. The use of pozzolana in a mortar provides the underwater curing (hydraulic limes) of whatever construction with higher speed compared to carbonation processes of slaked lime. Whenever pozzolana is not available, it is substituted by ceramic fragments, which possess similar hydraulic properties. This research focuses, for the first time, on the detailed characterization of mortars coming from the Thermal Complex of Baia, which represents one of the most important archaeological sites in the Campania region. Thanks to several thermal springs, the ancient city of Baiae (Campi Flegrei) was the holiday resort of the Roman aristocracy. The former Soprintendenza Archeologia della Campania, allowed us to perform non-invasive, but representative, sampling of mortars that were characterised by multianalytical methodologies (POM, XRPD, SEM-EDS, TGA, and MIP) providing useful information on possible future activities of restoration. Results confirmed the expertise of Roman workers, who skilfully combined volcanic tuff aggregate, hydrated lime, and ceramic fragments. In particular, the typical zeolitic mineral association of phillipsite > chabazite > analcime found in the tuff aggregate pointed out their provenance from the Neapolitan Yellow Tuff related to the volcanic activity of Campi Flegrei of ca. 15 ka BP. The most relevant characteristic detected in all studied samples is the mortar hydraulicity testified by evidences such as reaction rims between pozzolana and binder, Hydraulicity Index (HI), and thermal analyses investigation. Also, composition of secondary mineralogical phases in the cementiceous matrix is particularly relevant. Distinctive is the contemporary presence of C-A-S-H gel, calcite and gypsum. C-A-S-H gel is derived from lime/ceramic fragments reaction; calcite is likely related to the partial reaction of underburned lime; and gypsum could be ascribable to the sulphation process of calcite. These secondary minerogenetic products fill pore space and enhance bonding in pumice fragments, thus contributing to long-term durability of mortars. © 2019 Elsevier Masson SAS. All rights reserved.

1. Introduction and research aims

夽 This article is part of the special issue “Geosciences for Cultural Heritage”, composed of a selection of peer-reviewed papers presented at session S30 of Congress SGI-SIMP 2018. Guest editors: Fabrizio Antonelli (University IUAV of Venice), Alberto De Bonis (University of Naples “Federico II”), Domenico Miriello (University of Calabria), Simona Raneri (University of Pisa), and Alberta Silvestri (University of Padua). ∗ Corresponding author. E-mail address: [email protected] (C. Rispoli).

Whenever one thinks of Roman Empire, the mind goes straight to history and literature and not to geological science. However, geology should be definitely taken into account as the great technological progresses of this period were achieved through a skilful and intensive use of available georesources never seen until then. The ability to building roads, aqueducts, temples and monuments was so technologically developed that these manufacts still resist over two thousand years to the strength of subaerial weathering, waves and seawater chemical interactions. Roman craftsmen

https://doi.org/10.1016/j.culher.2019.05.010 1296-2074/© 2019 Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: C. Rispoli, et al., The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy), Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.05.010

G Model CULHER-3605; No. of Pages 12 2

ARTICLE IN PRESS C. Rispoli et al. / Journal of Cultural Heritage xxx (2019) xxx–xxx

Fig. 1. a: satellite picture of Terme di Baia; b: geological sketch map of Campi Flegrei [modified after 11].

Fig. 2. a: planimetry of Archaeological Park of Terme di Baia, along with samples location (modified after [18]); b: Venere area; c: Piccole Terme area; d: Sosandra area; e: Mercurio area; f: Villa dell’Ambulatio area. SV: Venere area samples; TM and V1: Mercurio area samples; PT: Piccole Terme area samples; VA: Villa dell’Ambulatio samples.

knew that thanks to the combination of lime with specific volcanic products (pozzolana), mortar and concrete become hydraulic allowing underwater hardening and increasing their mechanical strength [1]. The use of pozzolana marked a revolutionary progress in the construction sector, due to the ability of mixture to cure also underwater (hydraulic limes) and with a higher speed compared to carbonation processes of slaked lime. Whenever volcanic material was not available, fragments of artificial materials (ceramic fragments) with similar hydraulic properties were used [1]. The Department of Earth Sciences, Environment and Resources (DiSTAR) of the Federico II University of Naples, for over twenty years has been engaged in the application mineralogical and petrographic methods for archaeometric studies of several ancient finds and monuments, such as Roman ceramics, mortars and concrete [2–9].

Aim of this study is improving knowledge of Roman construction techniques used for the production of mortar-based materials from one of the most important archaeological sites of the Campania region of Italy: the archaeological park of Terme di Baia. The investigation was carried out by means of mineralogical, petrographic and physical techniques to examine in detail microstructural and compositional features of mortars and pointing out: mix design, provenance of raw materials, study of secondary minerogenetic processes. Moreover, the outcomes of this research will also represent a valuable base of information for future activities of restoration of this important archaeological site. 2. Geological and archaeological settings The archaeological site of Terme di Baia (Fig. 1a) is located in the western sector of Campi Flegrei volcanic district (Campania

Please cite this article in press as: C. Rispoli, et al., The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy), Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.05.010

G Model CULHER-3605; No. of Pages 12

ARTICLE IN PRESS C. Rispoli et al. / Journal of Cultural Heritage xxx (2019) xxx–xxx

region, southern Italy; [10,11] (Fig. 1b), representing with SommaVesuvius one of the two most important and still active quaternary volcanoes in the Mediterranean area. The Campi Flegrei is a volcanic field that caused different explosive events, which can produce huge volumes of weakly silica undersaturated trachytic-phonolitic pyroclastic rocks [10,12]. Volcanic activity in the area started around 80 ka [13] and was characterised by two main calderaforming eruptions, the Campanian Ignimbrite (39 ka) [14–16] and the Neapolitan Yellow Tuff (15 ka; Fig. 1b) [13,17]. The Archaeological Park of Terme di Baia covers an area of 40,000 m2 (Fig. 2) including ruins of outstanding patrician residences and Spa facilities dating back to the 2nd century BC. The entire complex consists in five areas (Fig. 2), built in different periods, from the 2nd centruy BC to the 4th century AD, and organized on different levels of terracing communicating with stepped ramps. Areas can be summarised as follows: the Venere area (Fig. 2b), Piccole Terme (Fig. 2c), the area of Sosandra (Fig. 2d), the area of Mercurio (Fig. 2e), and Villa dell’Ambulatio also known as complex of Terraces (Fig. 2f) [18]. The Venere area owns its name to eighteenth-century scholars who defined “Rooms of Venere” some rooms of the lower level of the complex, characterised by refined decorations on the vaults. It includes three buildings of different periods (2nd c. BC - 4th c. AD). The intermediate level of Venere area, called Piccole Terme consists of a circular-shaped laconicum and a pool, originally part of a late-republican villa and then integrated with other thermal environments such as the calidarium and tepidarium, when the building assumed a public function [19]. The nucleus, named Sosandra Area, occupies the central part of the entire complex. The structure is articulated on four levels: the residential part develops on the two upper levels. From the largest room in the lowest sector comes the marble statue of the Aspasia,

3

known as Aphrodite Sosandra (Roman copy of a Greek original), which gives its name to the sector. The two lower levels of the building are articulated with a scenographic effect and are constituted by a hemicycle and an uncovered area [19]. The so-called area of Mercurio takes its name from a thermal plant with a circular plan with domed vault called “Temple of Mercurio”. The dome cover of this temple anticipates the vault of the Pantheon in Rome an entire century earlier [19]. Finally, the area, called Villa dell’Ambulatio, is divided into six terraces with different functions, such as the domestic quarters and hospitalia, namely the accommodation for people using the nearby baths in the lower terrace of this area. Between the third and fifth century BC the phenomenon of bradyseism caused the slow sinking of piers, villas and buildings and contributed to the depopulation of the area, but the fame of its waters remained unchanged and its thermal plants survived until the 17th century [20]. 3. Materials and methods Due to the complexity and importance of the archaeological site, sampling was made in cooperation with responsible archaeologists and under the supervision of the former Soprintendenza Archeologia della Campania, now Parco Archeologico dei Campi Flegrei, which also provided the necessary authorisation. After a detailed inspection and a photographic survey of the different building materials characterising the site, we decided to focus the study on coating mortars, as they have been in contact with thermal water for a long time. A total of 21 samples, representative of the four sectors (Mercurio area, Venere area, Piccole Terme area and Villa dell’Ambulatio area) of the complex and ranging from the end 2nd century B.C. to 4th century A.D. (Fig. 2a), were collected following

Fig. 3. a–c: macroscopic images of mortar samples; d–m: microphotographs of mortar components (in CPL: Cross Polarized Light; PPL: Plane Polarized Light): d: cryptocrystalline matrix and plagioclase in PT1 sample (CPL); e: micritic matrix and clinopyroxene in SV5 sample (CPL); f: lime lump in TM2 sample (CPL); g: volcanic fragment in SV1 sample (CPL); h: pumice with reaction rim in TM3 sample (PPL); i: sanidine and amphibole in VA4 sample (CPL); l: ceramic fragment with high optical activity in TM3 sample (CPL); m: ceramic fragments with low optical activity in TM3 sample (CPL).

Please cite this article in press as: C. Rispoli, et al., The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy), Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.05.010

G Model CULHER-3605; No. of Pages 12 4

ARTICLE IN PRESS C. Rispoli et al. / Journal of Cultural Heritage xxx (2019) xxx–xxx

restrictive criteria aiming to both very limited invasiveness (i.e., small sample sizes, low visual impact) and high representativeness. The Sosandra sector was the only area where, for safety reasons, no sampling was allowed. In addition, detailed archive researches at Soprintendenza offices were performed in order to ensure that collected samples belonged to the original geomaterials, i.e., discarding those subjected to restoration works. The collected samples were characterised from a mineralogical and petrographic point of view according to the UNI-EN 11305 [21] and UNI-EN 11176 [22]. Macroscopic observation was the first approach for identification of the materials and for planning the analytical procedures, which were performed at the DiSTAR laboratories and at Group Technical Center (CTG) of Italcementi Heidelberg Group in Bergamo (Italy). The petrographic observation was carried out on thin sections through polarized optical microscopy (POM) with a Leica DFC280 microscope. The percentage of binder and aggregate was measured via modal analysis on four representative thin sections selected based on macro- and microscopic features of samples from each different sector of the investigated area. 1500 points for each section were counted using a Leica Q Win image analysis software. This analysis can be considered representative since the maximum uncertainty of percentage for a total amount of 1500 points is about 2.8% [23]. The qualitative mineralogical composition of the samples was obtained by X-Ray Powder Diffraction (XRPD) using a Panalytical X’Pert Pro diffractometer equipped with a RTMS X’Celerator detector with Cu-K␣ radiation, operating at 40 kV and 40 mA. Scans were collected in the range 4–70◦ 2␪ using a step interval of 0.017◦ 2␪, with a step counting time of 120 seconds. To identify the mineral phases in each X-ray powder spectrum, the Panalytical Highscore Plus 3.0c software and PDF-2/ICSD databases were used. Micro-textural observations and quantitative micro-chemical analyses were carried out by Scanning Electron Microscopy coupled with Energy Dispersive Spectroscopy (SEM/EDS; JEOL JSM-5310 coupled with Oxford Instruments Microanalysis Unit equipped with an INCA X-act detector). Measurements were performed with an INCA X-stream pulse processor using a 15-kV primary beam voltage, 50-100 ␮A filament current, variable spot size, from 30,000 to 200,000 × magnification, 20 mm WD and 50 s net acquisition real time). The INCA Energy software was employed, using the XPP matrix correction scheme, developed by Pouchou & Pichoir [24], and the Pulse Pile up correction. The quant optimization was carried out using cobalt (FWHM–full width at half maximum peak height- of the strobed zero = 60–65 eV). The following Smithsonian Institute and MAC (Micro-Analysis Consultants Ltd) standards (see ESM1) were used for calibration: diopside (Ca), fayalite (Fe), San Carlos olivine (Mg), anorthoclase (Na, Al, Si), rutile (Ti), serandite (Mn), microcline (K), apatite (P), fluorite (F), pirite (S) and sodium chloride (Cl). Ka lines were used for calibration of all elements; detection limits of the analysed elements are below 0.1%. Precision and accuracy of EDS analyses has been calculated on the basis of multiple acquisition (10 point data) of Laki 1783 natural glass [25]. Precision is below 1% for SiO2 , ranging from 1 to 6% for Na2 O, MgO, Al2 O3 , CaO, TiO2 , and major of 10% for the other elements (see ESM1); accuracy is lower than 1% for, MgO, Al2 O3 , SiO2 , CaO, MnO, FeO and P2 O5 , ranging from 2% to 10% for other elements (see ESM1). Micro-chemical analyses were performed to determine major chemical composition of the binder and the lime lumps of the samples and to study the volcanic fragments present among the aggregates. The hydraulicity index (HI) of binder was calculated according to the method of Boynton [26] using the (SiO2 + Al2 O3 + Fe2 O3 )/(CaO + MgO) ratio. It was measured on thin sections by SEM-EDS analyses on areas of homogeneous aspect (10 ␮m spot-size). Since the value of the HI can be influenced by presence of microfragments of volcanic components scattered in

Table 1 Modal analyses of mortars. Constituents (%)

Mortars PT2 PT1

TM2

SV1

VA1

Feldspars (Afs, Pl) Mafic minerals (Cpx, Am, Bt) Volcanic fragments Scoriae Pumice Ceramic fragments Carbonate fragments Sparite Lime lumps Micritic matrix Cryptocristalline matrix Voids Others Total points Total binder Total sggregate Binder/Aggregate ratio

4.7 2.1 0.9 16.6 15.6 9.7 0.5 1.1 3.3 12.3 30.9 1.9 0.5 100.0 47.6 50.0 1.0

7.5 1.6 1.0 11.5 5.5 20.3 0.5 1.2 4.2 5.5 35.1 6.3 0.0 100.0 46.0 47.6 1.0

2.8 1.3 1.3 4.1 11.8 6.7 0.4 1.4 1.5 11.0 51.4 6.1 0.0 100.0 65.3 28.5 2.3

6.4 1.3 1.3 9.5 14.3 11.7 0.0 1.9 1.3 28.6 16.8 7.0 0.0 100.0 48.6 44.5 1.1

4.6 1.6 3.8 4.2 8.9 7.3 0.5 3.7 2.5 10.6 47.3 5.0 0.0 100.0 64.1 30.9 2.1

Abbreviations: Afs: alkali-feldspar; Pl: plagioclase; Cpx: clinopyroxene; Am; amphibole; Bt: Biotite.

the binder, it must be considered as a semiquantitative measure to obtain preliminary information about the binder’s composition [27]. Thermal Analyses (TGA/SDTA) were carried out with a Mettler Toledo TGA/SDTA 851e instrument and Mettler Toledo STARe SW 7.01 software, with the main goal of determining the total (binder plus aggregates) hydraulic features of these materials. The weight loss of mass of 20–50 mg was monitored through a heating cycle from 25 to 1000 ◦ C with a temperature gradient of 10 ◦ C/minute in nitrogen atmosphere (flow 60 mL/min). This process reveals thermal transformations such as dehydration, dihydroxylation, oxidation and decomposition, giving quantitative information on binder compounds. TGA data were processed according to Moropoulou et al. [28,29] in the range 200–600 ◦ C and weight loss due to water bound to hydraulic components was determined; for temperatures > 600 ◦ C, weight loss due to decomposition of carbonates was obtained [30,31]. The pore system of samples was investigated by means of Mercury Intrusion Porosimetry (MIP). Due to the scarce amount and small dimensions of samples, on the bases of macroscopic and microscopic features four fragments were selected, approximately 1 cm3 in size. The selected fragments were dried in an oven for 24 h at 105 ◦ C, and then analysed on Thermo Scientific equipments PASCAL 140 with a maximum injection pressure of 0.4 Mpa and PASCAL 240 with a maximum injection pressure of 200 MPa. These instruments used consecutively provided: total volume of pores of radius between 3.75 nm and 800 ␮m (expressed in mm3 /g), open porosity (expressed in vol %), bulk density (expressed in g/cm3 ), apparent density (expressed in g/cm3 ) and specific surface (expressed in m2 /g), graphical and numerical representation of the distribution of pore size. 4. Results 4.1. Texture and optical microscopy of the mortar-based materials Macroscopic observation revealed a high similarity among coating mortars (Fig. 3a–c). They display a high degree of cohesion, except for VA1 sample that is friable. The samples show a variable colour ranging from light yellow to brownish (MUNSELL 2.5YR 6/3 to 10YR 4/4 colours) [32]. Ceramic and volcanic fragments mainly represent visible aggregates; in some samples carbonate fragments are also visible. The grain size of aggregates varies from 1 to 40 mm.

Please cite this article in press as: C. Rispoli, et al., The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy), Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.05.010

G Model CULHER-3605; No. of Pages 12

ARTICLE IN PRESS C. Rispoli et al. / Journal of Cultural Heritage xxx (2019) xxx–xxx

Petrographic analysis confirmed a high homogeneity in the collected mortars (Fig. 3d–m, Table 1). They are characterised by a binder, yellowish to brownish colour, and cryptocrystalline (16.8–51.4%; Fig. 3d) to micritic texture (5.5–28.6%, Fig. 3e). The binder shows the presence of lime lumps that represent unreacted lime [33,34], with undefined and irregular edges (1.3–4.2%, Fig. 3f), and sparite grains (1.1–3.7%). Aggregates are represented by volcanic fragments (0.9–3.8%; Fig. 3g), pumice (5.5–15.6%; Fig. 3h), and scoriae (4.1–16.6%) with evident reaction rims, and by crystal fragments (7.5%) of sanidine, amphibole (Fig. 3i), clinopyroxene, biotite and plagioclase (Fig. 3d; feldspars: 2.8–7.5%; mafic minerals: 1.3–2.1%) along with ceramic fragments (6.7–20.3%, Fig. 3l and m). Some samples revealed also the presence of carbonate fragments (0.5%). The identified volcanic fragments can be ascribed to a volcanic tuff, due to the presence of microcrystals dispersed in an ashy matrix [35]. Ceramic fragments are different from each other, even if included in a single mortar sample. In particular, ceramic matrices can be either characterised by high (Fig. 3l) or low (Fig. 3m) optical activity. In some fragments inclusions are mainly represented by tiny crystals of quartz and alkali-feldspar, whereas in other fragments they are characterised by a prevailing volcanic components (clinopyroxene, volcanic glass, pumice, scoriae, and sanidine). Modal analysis (Table 1) carried out on representative samples (PT1, PT2, TM2, SV1, and VA1), highlighted that void percentage ranges from 1.9 to 7.0% and the binder/aggregate ratio reaches values higher than 2 in PT2 and SV1 samples; the other mortars (PT1, TM2 and VA1) show a binder/aggregate ratio equal to 1. 4.2. Mineralogical analysis The XRPD analysis was performed on three different fractions (binder, aggregate and ceramic fragments) separated from mortar samples according to UNI-EN 11305 [21], and their results are summarised in ESM 2. The main binder phase in all samples is calcite; gypsum is detected in only 9 samples (PT3, PT5, SV2, SV3, SV4, SV5, TM1, TM5and VA3) and aragonite in sample V1. The mineralogical phases identified in the aggregates are phillipsite, analcime, chabazite, sanidine, clinopyroxene and mica. The main mineralogical phases in the ceramic fragments are quartz, calcite along with sanidine, clinopyroxene, plagioclase, mica, and subordinate hematite. All analysed samples are characterised by the presence of halite.

5

The sample with the highest HI (0.38) value is V1, thus confirming its difference from all other samples. Additional information was obtained by EDS microanalyses on pumice fragments (Table 3). They are classified as trachytes according to the Total Alkali versus Silica diagram (TAS; Fig. 6) [41]. 4.4. Thermogravimetric analysis Thermal analyses showed that studied mortars have a percentage of structural bound water (SBW, the amount of OH groups inside the silicatic framework and inside the thinnest ultramicropores whose diameter is comparable to that of a water molecule) and CO2 (derived from the calcination of carbonatic phases) content that allowed us to classify them as hydraulic mortars (CO2 /SBW < 3%; Table 4) [29]. The highest SBW content is reported for sample PT2 (13.63%) and the lowest for sample TM3 (5.23%), but this latter sample displays also the lowest total LOI (Loss On Ignition; 20.41%). CO2 values range from 4.31% (sample VA1) to 12.33% (sample TM1), except for sample V1, which shows a very high value of CO2 content (20.56%). In TM samples, CO2 values are always higher than SBW content, with the only exception of sample TM5. According to literature data, using Moropoulou et al. [29] method, the CO2 /SBW ratio of mortars can be defined as naturalpozzolanic mortars (Fig. 7) whereas sample V1 belongs to artificial pozzolanic mortars. 4.5. Mercury Intrusion Porosimetry (MIP) Porosity was determined on pieces of selected mortar (PT2, TM3, TM4, and VA2) due to the scarce amount of available material. Table 5 shows the parameters of MIP (cumulative volume, bulk density, open porosity, specific surface and apparent density) determined for the four samples; whereas Fig. 8 shows the representative pore size distribution. Relative volume curves of investigated samples are positively skewed and display a pore radii mainly ranging between 5 and 100 nm, except for TM3 where a displacement of the pore access radius towards larger sizes (100–1000 nm) is recognised (Fig. 8). The open porosity ranges from 38.2 to 49.9 Vol. (%) with unimodal and broadened shape of the cumulative pore size distribution (Fig. 8). In particular, three samples (TM3, TM4 and VA2) show quite similar values of the open porosity, while PT2 shows the highest value (Table 5).

4.3. Microstructural and microchemical analyses

5. Discussion and conclusions

SEM analyses, performed on the binder fragments, confirmed the presence of gypsum in some samples (Table 2; Fig. 4a) and demonstrated the presence of newly-formed hydraulic phases like C-A-S-H gel (Calcium Aluminum Silicate Hydrate; Fig. 4b), with their typical spongy morphology [36–38]. The SEM analyses on volcanic fragments revealed the presence of phillipsite crystals, with their prismatic habit, and pseudocubic crystals of chabazite (Fig. 4c), [33,39,40], mica, with lamellar habit (Fig. 4d) and smectite showing “flakes” structure (Fig. 4e). Volcanic glass is occasionally altered (Fig. 4f). EDS microanalyses carried out on binder and lime lumps of the mortars (Table 2) provided us chemical information on their composition. These results revealed that lime lumps mainly consisted of CaO showing very high values of CaO + MgO comprised between 88.79 and 96.35 wt.%, compared with binder (70.83–85.03 wt.%). The binder is also characterised by concentrations of SiO2 + Al2 O3 + Fe2 O3 (10.17–26.94 wt.%) higher than in lime lumps (2.71–7.49 wt.%) (Table 2). The HI of lime lumps is lower than 0.10, while the HI of binder ranges between 0.12 and 0.38 (Fig. 5).

This is the first study performed with a multianalytical approach focused on mortar-based materials from the Terme di Baia complex, which represents an outstanding example of the Roman construction techniques. Provenance and typology of raw materials along with structural and textural features related to binder and aggregates were defined. The mineralogical, petrographic and compositional data confirmed that these coating mortars were made by mixing slaked lime, water, and volcanic, ceramic and carbonate aggregates, with fine and/or coarse grain size. This typical mix design was described, for the first time, by Vitruvius in De Architectura (Liber VIII), who called it signinum, a name for indicating the waterproof mortars used in water tanks, in thermal pools and in the caverns of aqueducts. Considering the surrounding geological setting, volcanic aggregates have a local provenance. The presence of phillipsite, analcime, and chabazite, revealed by XRPD and SEM-EDS analyses, represents the typical zeolitic association found in Phlegraean volcanic deposits and, precisely, in the Neapolitan Yellow Tuff (NYT), one of the most important and widespread eruptive products deriving

Please cite this article in press as: C. Rispoli, et al., The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy), Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.05.010

VA2 L

VA2 B

VA3 L

VA3 B

VA4 L

VA4 B

VA5 L

VA5 B

SV1 L

SV1 B

SV2 L

SV2 B

SV3 L

SV3 B

SV4 L

SV4 B

SV5 L

SV5 B

PT1 L

11.19
4.39
10.83
3.99
11.10
3.80
10.21
4.03
11.07
4.78 0.11 0.91
8.59
3.34
11.18 0.20 2.72 0.52 0.50 5.16 75.17 1.63 0.66
4.18 0.14 0.84
10.41
3.98
10.75
3.28
11.02
5.16
SiO2 +Al2 O3 + Fe2 O3 CaO + MgO HI

6.28

15.83

5.61

15.59

5.29

15.59

4.95

14.43

5.76

15.53

5.69

12.67

4.44

14.41

5.02

14.27

4.93

14.75

4.27

15.31

6.48

89.42 0.07

78.93 0.20

90.13 0.06

79.88 0.20

90.61 0.06

80.64 0.19

91.19 0.05

81.44 0.18

90.69 0.06

79.89 0.19

91.62 0.06

83.13 0.15

93.69 0.05

80.34 0.18

92.63 0.05

81.91 0.17

93.62 0.05

81.29 0.18

94.65 0.05

81.93 0.19

89.45 0.07

wt. %

PT1 B

PT2 L

PT2 B

PT3 L

PT3 B

PT4 L

PT4 B

PT5 L

PT5 B

TM1 L

TM1 B

TM2 L

TM2 B

TM3 L

TM3 B

TM4 L

TM4 B

TM5 L

TM5 B

V1 L

V1 B

SiO2 TiO2 Al2 O3 Fe2 O3 MnO MgO CaO Na2 O K2 O P2 O5 SO3 FClTotal

17.42 0.24 4.46
6.08 0.22 1.78 0.22
11.00
5.05
12.50
4.03
10.47
3.43
11.15
5.49 0.27 1.09 0.11
8.32 0.26 1.85
6.18
14.76 0.39 3.02 0.12
1.46 0.13 0.93
12.87
3.22
12.09
4.03
13.05
1.93
17.82 0.18 8.04 1.09
SiO2 +Al2 O3 + Fe2 O3 CaO + MgO HI

21.88

8.08

14.94

6.49

15.32

5.24

14.09

4.76

15.35

6.69

10.17

7.42

17.90

2.39

16.47

4.29

15.54

5.46

19.02

2.71

26.95

72.47 0.30

91.05 0.09

80.10 0.19

91.91 0.07

81.21 0.19

91.40 0.06

80.61 0.18

91.91 0.05

80.37 0.19

89.02 0.08

85.03 0.12

88.79 0.08

79.27 0.23

96.08 0.02

79.01 0.21

92.09 0.05

81.90 0.19

92.63 0.06

78.72 0.24

96.35 0.03

70.83 0.38

ARTICLE IN PRESS

VA1 B

4.87
G Model

VA1 L

SiO2 TiO2 Al2 O3 Fe2 O3 MnO MgO CaO Na2 O K2 O P2 O5 SO3 FClTotal

C. Rispoli et al. / Journal of Cultural Heritage xxx (2019) xxx–xxx

wt. %

CULHER-3605; No. of Pages 12

6

Please cite this article in press as: C. Rispoli, et al., The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy), Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.05.010

Table 2 Major element concentrations of lime lumps (L) and binder (B) in the Terme di Baia mortars (in wt.%; recalculated to 100%; D.L.: Detection Limit= 0.1 wt. %). SiO2 + Al2 O3 + Fe2 O3 , CaO + MgO, HI (Hydraulic Index) also shown. Note: Fe2 O3 = FeO *1.11.

G Model CULHER-3605; No. of Pages 12

ARTICLE IN PRESS C. Rispoli et al. / Journal of Cultural Heritage xxx (2019) xxx–xxx

7

Fig. 4. SEM images of: a: gypsum in SV2 binder; b: C-A-S-H gel in TM3 binder; c: phillipsite and chabazite in PT2 volcanic fragment; d: mica in VA1 volcanic fragment; e: smectite in PT1 volcanic fragment; f: altered volcanic glass in TM1 fragment.

Fig. 5. Hydraulicity Index (HI) values: lime lumps (black) and binder (grey) of mortar samples.

Fig. 6. Classification of pumice fragments in the Terme di Baia mortars compared with Campi Flegrei volcanic glasses (CI and NYT; [10]), using TAS diagram [41]. Abbreviations: CI: Campanian Ignimbrite; NYT: Neapolitan Yellow Tuff; TB: Terme di Baia pumice fragments.

Please cite this article in press as: C. Rispoli, et al., The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy), Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.05.010

VA3

VA4

VA5

SV1

SV2

SV3

SV4

SV5

V1

SiO2 TiO2 Al2 O3 Fe2 O3 MnO MgO CaO Na2 O K2 O P2 O5 SO3 FCl Total

63.00
60.65 0.16 18.41 3.30 0.39 0.35 2.07 6.05 7.12 0.11 0.19 0.59 0.61 100.00

61.27 0.53 18.13 3.20 0.12 0.36 1.84 5.38 8.09 0.22
61.69 0.51 18.22 2.66 0.21 0.23 1.80 5.50 8.10
56.35 0.69 18.62 5.41 0.27 1.33 4.40 3.62 8.24 0.41 0.12
61.48 0.21 17.86 3.14 0.28 0.33 2.15 5.03 7.50
61.35 0.59 17.18 4.57 0.11 0.41 1.80 5.80 7.14
61.74 0.50 18.25 3.36 0.36 0.27 1.89 5.17 7.65
61.18 0.54 17.92 3.18 0.32 0.35 1.98 5.18 8.01 0.20 0.27
61.56 0.41 18.13 3.29
61.19 0.27 17.82 3.07 0.29 0.19 1.97 5.78 7.81
Na2 O + K2 O

13.27

13.17

13.47

13.60

11.86

12.52

12.94

12.83

13.19

13.24

13.59

wt.%

PT1

PT2

PT3

PT4

PT5

TM1

TM2

TM3

TM4

TM5

SiO2 TiO2 Al2 O3 Fe2 O3 MnO MgO CaO Na2 O K2 O P2 O5 SO3 FCl Total

61.89 0.30 17.97 3.07
61.61 0.47 16.67 4.74 0.17 0.51 1.89 5.59 6.96 0.15 0.15
61.21 0.46 17.31 4.49 0.20 0.30 2.17 5.86 7.09
61.48 0.48 17.85 2.98 0.45 0.26 2.12 5.76 7.88
61.54 0.03 18.81 2.80 0.30 0.22 2.30 6.39 6.91
60.83 0.37 16.60 4.63 0.70 0.48 2.15 6.59 6.52
62.01 0.42 18.16 3.18 0.46 0.21 1.92 5.44 7.11 0.25
60.73 0.67 17.18 4.17 0.45 0.56 2.19 5.74 6.94 0.25
61.93 0.41 18.31 2.58 0.25 0.24 1.67 6.05 7.19
61.70 0.45 18.20 2.87
Na2 O + K2 O

13.05

12.54

12.94

13.64

13.30

13.12

12.55

12.68

13.24

13.55

ARTICLE IN PRESS

VA2

G Model

VA1

C. Rispoli et al. / Journal of Cultural Heritage xxx (2019) xxx–xxx

wt.%

CULHER-3605; No. of Pages 12

8

Please cite this article in press as: C. Rispoli, et al., The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy), Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.05.010

Table 3 Major element concentrations of pumice fragments in the Terme di Baia mortars (in wt.%, recalculated to 100%; D.L.: Detection Limit = 0.1 wt. %). Na2 O + K2 O also shown. Note: Fe2 O3 = FeO *1.11.

G Model

ARTICLE IN PRESS

CULHER-3605; No. of Pages 12

C. Rispoli et al. / Journal of Cultural Heritage xxx (2019) xxx–xxx

9

Table 4 Thermal analysis features of investigated samples. Abbreviation: SBW: Structural Boundary Water; LOI: Loss On Ignition. Sample

SBW %

CO2 %

CO2 /SBW

LOI

PT1 T range (◦ C) PT2 T range (◦ C) PT3 T range (◦ C) PT4 T range (◦ C) PT5 T range (◦ C) SV1 T range (◦ C) SV2 T range (◦ C) SV3 T range (◦ C) SV4 T range (◦ C) SV5 T range (◦ C) TM1T range (◦ C) TM2T range (◦ C) TM3T range (◦ C) TM4T range (◦ C) TM5T range (◦ C) VA1T range (◦ C) VA2T range (◦ C) VA3T range (◦ C) VA4T range (◦ C) VA5T range (◦ C) V1 T range (◦ C)

8.54 240–650 13.63230–680 9.11 235–620 8.14 230–640 10.36235–610 7.12 200–570 9.42 200–560 7.74 210–570 8.79 200–580 9.35 200–570 8.64 200–650 10.05210–630 5.23 200–640 6.84 200–630 9.15 220–640 9.35 240–630 10.67240–650 8.80 230–620 6.94 240–650 9.78 220–640 3.38 210–560

4.79 650–750 6.24 680–760 7.57 650–740 5.69 660–750 7.23 650–750 6.34 640–760 7.11 560–760 5.33 570–750 6.26 580–740 5.77 570–750 12.33650–780 12.13630–780 11.23640–790 9.41 630–780 6.41 640–760 4.31 630–750 7.02 650–740 7.39 620–750 8.99 650–740 6.44 640–730 20.56560–850

0.56 0.46 0.83 0.70 0.70 0.89 0.75 0.69 0.71 0.62 1.43 1.20 2.15 1.38 0.70 0.46 0.66 0.84 1.30 0.66 6.08

28.3225–1000 24.4925–1000 31.1525–1000 29.1325–1000 26.3525–1000 27.4525–1000 33.1725–1000 24.2025–1000 27.3625–1000 30.2525–1000 32.4725–1000 35.2225–1000 20.4125–1000 30.8325–1000 29.4125–1000 25.5725–1000 29.5625–1000 25.1925–1000 31.0925–1000 28.9625–1000 30.0625–1000

Fig. 7. Binary CO2 /SBW vs. CO2 (%) diagram utilised to compare the obtained data from Terme di Baia mortars (TB samples; dark grey diamonds) and those of Moropoulou et al., [29]. Abbreviations: NPM: natural pozzolanic mortars; APM: artificial pozzolanic mortars; HLM: Hydraulic lime mortars; LM: lime mortars.

from the volcanic activity of the area, dating to 15ka BP [35,39]. This information was supported by the chemical composition of pumice fragments that, according to the TAS diagram follows the compositional trend defined by NYT volcanic glasses (Fig. 6). Regarding ceramic fragments, it was not possible to define with certainty their provenance, as the extreme petrographic differences likely suggest a recycling of different fictile materials. The calcareous raw material and subsequent neoformation phases provided useful information on the provenance and production technology. The lime utilised in cementiceous binding matrix and the carbonate fragments identified in the aggregates may be considered as a local as all the Campanian Plain is bordered by carbonate mountain ridges of Mesozoic age [10]. Aragonite was detected in V1 sample. The presence of this phase may suggest that the bathtub from which the sample was taken in the Mercurio area was likely built with a different limestone. The mineralogical composition of cementiceous binding matrix shows contemporary presence of calcite, gypsum and C-A-S-H

Table 5 Porosimetric features of Terme Di Baia mortars. Sample

PT2

TM3

TM4

VA2

Cumulative volume (mm3 /g) Bulk density (g/cm3 ) Open porosity (Vol. %) Apparent Density (g/cm3 )

370.7 1.35 49.9 2.70

244.5 1.58 38.2 2.55

295.3 1.42 39.8 2.37

295.6 1.46 42.5 2.56

gel, with traces of halite. Each of these newly-formed phases are strictly related to different processes. The calcite formation can be related to the not-well reacted clasts of under-burned lime. Alternatively, carbonation processes from residual slaked lime could have occurred in mortars, since they have been exposed in a subaerial environment. The gypsum, the main widespread newlyformed mineral, is related to calcite sulphation as a consequence of the decrease of pH value, caused by dissolution of atmospheric SO2 [42]. The presence of halite can be attributed to the interaction

Please cite this article in press as: C. Rispoli, et al., The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy), Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.05.010

G Model CULHER-3605; No. of Pages 12 10

ARTICLE IN PRESS C. Rispoli et al. / Journal of Cultural Heritage xxx (2019) xxx–xxx

Fig. 8. Cumulative (continuous line) and relative (dashed line) pore size distribution of Terme di Baia mortars (PT2, TM3, TM4 and VA2).

with seawater in ancient time or marine aerosol [43]. On the other hand, the C-A-S-H gel derived from reactions between lime and pozzolanic material (volcanic and ceramic materials); its formation testifies the mortar hydraulicity [44,45]. Another very important goal obtained in this research was the determination of hydraulicity of the Terme di Baia mortars, via SEMEDS and thermal analyses. The lime lumps with a HI lower than 0.10, can be classified as quicklime or aerial lime (Fig. 5) [46–48], while the HI of binders (0.12–0.38) categorise these geomaterials as weakly-moderately hydraulic materials. Thus, these HI values confirmed that mortar hydraulicity is strictly associated with the presence of pozzolanic materials (both ceramic and volcanic fragments), which increases considerably the reaction between silica, aluminum and calcium hydroxide leading to the neoformation of calcium aluminium silicate hydrates, the so-called C-A-S-H gel [45,49]. The hydraulicity of mortars was supported also by percentage of structurally bound water (SBW) and CO2 content obtained via thermal analyses (Table 4); in fact their thermal behaviour is related to the presence of hydrated compounds such as C-A-S-H gel, decomposition of calcite and other carbonates, that usually occur between 600–850 ◦ C with a consequent release of CO2 [50]. Sample V1 shows a very high percentage of CO2 (20.56%), thus confirming its difference from the other samples probably due to a different relative percentage of raw materials. Actually, this sample belongs to a bathtub stratigraphically more recent [19]. The secondary mineralogical phases, the C-A-S-H gel, play a crucial role for the durability of mortars, as they fill the pores and enhance bonding in pumice clasts, improving resistance to weathering [38,51]. This statement is supported by porosity tests that show for ancient roman mortars a maximum in pore radius distribution between 5 and 100 nm and a open porosity between 38–50 Vol.% (Fig. 8, Table 5). According to Sutter [52] this range is not considered detrimental for mortar durability. Only sample TM3, where the pore access radius goes towards larger sizes (100–1000 nm), could be more vulnerable to weathering [53–55]. The investigation performed on mortars, sampled in a diachronic sequence of about four centuries from different contexts of the Terme di Baia complex, highlighted the continuous use of a specific and valuable mix design created with local raw materials. Results of microchemical and thermal analyses evidenced that the use of local materials was of fundamental importance for producing hydraulic mortars. As a matter of fact, it is well known that

in this period the pulvis puteolanus (literally powder from Pozzuoli) was the main ingredient for improving the hydraulic properties of mortars and concrete, so that it was exported all over the Mediterranean area [8]. Moreover, the thermal complex of Baia is exactly located in the place identified by Vitruvius (De Architectura, Book II, Chapter 6) as the source of pozzolana: “est etiam genus pulveris, quod efficit naturaliter res admirandas. Nascitur in regionibus Bajanis, et in agris municipiorum, quae sunt circa Vesuvium montem, quod commixtum cum calce et caemento, non modo caeteris aedificiis praestat firmitates, sed etiam moles quae construuntur in mari, sub aqua solidescunt” (English translation: There is also a kind of powder which, by nature, produces wonderful results. It is found in the neighbourhood of Baiae and in the lands of the municipalities round Mount Vesuvius. This being mixed with lime and rubble, not only furnishes strength to other buildings, but also, when piers are built in the sea, they set under water). Considering that this thermal complex was designated for the Roman aristocracy, the choice had to be addressed to the best building material available in that time. Finally, this research can also represent a valuable reference on possible future activities of restoration in a very extended and important archaeological site.

Acknowledgements The authors would like to thank Dr. Roberto de Gennaro for invaluable assistance during EDS microanalyses, and Dr. Sergio Bravi for his technical ability in thin section preparation. Many thanks are due to the former Soprintendenza per i Beni Archeologici della Campania for courtesy of Dr. Pierfrancesco Talamo that allowed and guided the sampling phase. We also thank Prof. Paola Petrosino for providing all the necessary information regarding accuracy and precision of the EDS instrument. Thanks, are also due to CTG Italcementi Heidelberg Group for supporting this research activity. Finally, the authors wish to warmly thank the Editor and two anonymous reviewers for their useful suggestions and comments that definitely improved the manuscript.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.culher.2019.05.010.

Please cite this article in press as: C. Rispoli, et al., The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy), Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.05.010

G Model CULHER-3605; No. of Pages 12

ARTICLE IN PRESS C. Rispoli et al. / Journal of Cultural Heritage xxx (2019) xxx–xxx

References [1] M. Collepardi, La lezione dei romani: durabilità e sostenibilità delle opere architettoniche e strutturali, in: III Convegno Aim. “Restauro e Conserv. Dei Beni Cult. Mater. e Tec., Cassino (FR), 2003. [2] C. Germinario, G. Cultrone, L. Cavassa, A. De Bonis, F. Izzo, A. Langella, M. Mercurio, V. Morra, P. Munzi, C. Grifa, Local production and imitations of Late Roman pottery from a well in the Roman necropolis of Cuma in Naples, Italy, Geoarchaeology 34 (2019) 62–79, http://dx.doi.org/10.1002/gea.21703. [3] A. De Bonis, I. Arienzo, M. D’Antonio, L. Franciosi, C. Germinario, C. Grifa, V. Guarino, A. Langella, V. Morra, Sr-Nd isotopic fingerprinting as a tool for ceramic provenance: Its application on raw materials, ceramic replicas and ancient pottery, J. Archaeol. Sci. 94 (2018) 51–59, http://dx.doi.org/10.1016/j.jas.2018.04.002. [4] S.F. Graziano, C. Di Benedetto, V. Guarino, C. Rispoli, P. Munzi, P. Cappelletti, V. Morra, Technology and building materials in Roman age (1st BC - 2nd AD): The “Mausoleo Della Sfinge” from the archaeological site of Cuma (Italy), Mediterr. Archaeol. Archaeom. 18 (2018) 81–94, http://dx.doi.org/10.5281/zenodo.1256057. [5] C. Di Benedetto, S.F. Graziano, V. Guarino, C. Rispoli, P. Munzi, V. Morra, P. Cappelletti, Romans’ established skills: mortars from D46b Mausoleum, Porta Mediana Necropolis, Cuma (Naples), Mediterr. Archaeol. Archaeom. 18 (2018) 131–146, http://dx.doi.org/10.5281/zenodo.1285895. [6] V. Guarino, A. De Bonis, I. Faga, D. Giampaola, C. Grifa, A. Langella, V. Liuzza, R. Pierobon Benoit, P. Romano, V. Morra, Production and circulation of thin walled pottery from the Roman port of Neapolis, Campania (Italy), Period. Di Mineral. 85 (2016) 95–114, http://dx.doi.org/10.2451/2016PM618. [7] F. Izzo, A. Arizzi, P. Cappelletti, G. Cultrone, A. De Bonis, C. Germinario, S.F. Graziano, C. Grifa, V. Guarino, M. Mercurio, V. Morra, A. Langella, The art of building in the Roman period (89 B.C. - 79 A.D.): Mortars, plasters and mosaic floors from ancient Stabiae (Naples, Italy), Constr. Build. Mater. 117 (2016) 129–143, http://dx.doi.org/10.1016/j.conbuildmat.2016.04.101. [8] C. Stanislao, C. Rispoli, G. Vola, P. Cappelletti, V. Morra, M. De Gennaro, Contribution to the knowledge of ancient Roman seawater concretes: Phlegrean pozzolan adopted in the construction of the harbour at Soli-Pompeiopolis (Mersin, Turkey), Period. Di Mineral. 80 (2011), http://dx.doi.org/10.2451/2011PM0031. [9] C. Grifa, V. Morra, A. Langella, P. Munzi, Byzantine ceramic production from Cuma (Campi Flegrei, Napoli), Archaeometry 51 (2009) 75–94, http://dx.doi.org/10.1111/j.1475-4754.2008.00416.x. [10] V. Morra, D. Calcaterra, P. Cappelletti, A. Colella, L. Fedele, R. de Gennaro, A. Langella, M. Mercurio, M. de Gennaro, Urban geology: Relationships between geological setting and architectural heritage of the Neapolitan area, J. Virtual Explor. 36 (2010), http://dx.doi.org/10.3809/jvirtex.2010.00261. [11] L. Fedele, D.D. Insinga, A.T. Calvert, V. Morra, A. Perrotta, C. Scarpati, 40 Ar/39 Ar dating of tuff vents in the Campi Flegrei caldera (southern Italy): Toward a new chronostratigraphic reconstruction of the Holocene volcanic activity, Bull. Volcanol. 73 (2011) 1323–1336, http://dx.doi.org/10.1007/s00445-011-0478-8. [12] C. Rispoli, L. Fedele, C. Di Benedetto, R. Esposito, S.F. Graziano, V. Guarino, V. Morra, P. Cappelletti, Characterization of building materials from the Anfiteatro Flavio (Pozzuoli, southern Italy): A mineralogical and petrographic study, Ital, J. Geosci. 138 (2019) 103–115, http://dx.doi.org/10.3301/IJG.2018.29. [13] C. Scarpati, P. Cole, A. Perrotta, The Neapolitan Yellow Tuff - A large volume multiphase eruption from Campi Flegrei, Southern Italy, Bull. Volcanol. 55 (1993) 343–356, http://dx.doi.org/10.1007/BF00301145. [14] A. Langella, D.L. Bish, P. Cappelletti, G. Cerri, A. Colella, R. de Gennaro, S.F. Graziano, A. Perrotta, C. Scarpati, M. de Gennaro, New insights into the mineralogical facies distribution of Campanian Ignimbrite, a relevant Italian industrial material, Appl. Clay Sci. 72 (2013) 55–73, http://dx.doi.org/10.1016/j.clay.2013.01.008. [15] L. Fedele, C. Scarpati, M. Lanphere, L. Melluso, V. Morra, A. Perrotta, G. Ricci, The Breccia Museo formation, Campi Flegrei, southern Italy: Geochronology, chemostratigraphy and relationship with the Campanian Ignimbrite eruption, Bull. Volcanol. 70 (2008) 1189–1219, http://dx.doi.org/10.1007/s00445-008-0197-y. [16] L. Fedele, C. Scarpati, D. Sparice, A. Perrotta, F. Laiena, A chemostratigraphic study of the Campanian Ignimbrite eruption (Campi Flegrei, Italy): Insights on magma chamber withdrawal and deposit accumulation as revealed by compositionally zoned stratigraphic and facies framework, J. Volcanol. Geotherm. Res. 324 (2016) 105–117, http://dx.doi.org/10.1016/j.jvolgeores.2016.05.019. [17] A.L. Deino, G. Orsi, S. de Vita, M. Piochi, The age of the Neapolitan Yellow Tuff caldera-forming eruption (Campi Flegrei caldera - Italy) assessed by 40 Ar/39 Ar dating method, J. Volcanol. Geotherm. Res. 133 (2004) 157–170, http://dx.doi.org/10.1016/S0377-0273(03)00396-2. [18] R. Di Bonito, R. Giamminelli, Le Terme dei Campi Flegrei, Jandi Sapi Ed, Roma, 1992. [19] P. Amalfitano, G. Comodeca, M. Medri, I Campi Flegrei, un itinerario archeologico, Marsilio ed, Venezia, 1990 (ISBN: 8831753541). [20] M. Borriello, A. D’ambrosio, Baiae-Misenum, Forma Italiae, Regio I. Vol. XIV, 1979 (ISBN: 9788822228109). [21] UNI-EN 1130520 09, Beni culturali: Malte storiche, linee guida per la caratterizzazione mineralogico petrografica, fisica e chimica delle malte, 2009. [22] UNI-EN 1117620 06, Beni culturali - Descrizione petrografica di una malta, 2006. [23] R.J. Howarth, Improved estimators of uncertainty in proportions, pointcounting, and pass-fail test results, Am. J. Sci. 298 (1998) 594–607.

11

[24] J.L. Pouchou, F. Pichoir, Simplified version of the PAP model for matrix corrections in EPMA, Newbury DE, San Francisco, 1988. [25] S.C. Kuehn, D.G. Froese, P.A.R. Shane, INTAV Intercomparison Participants, The INTAV intercomparison of electron-beam microanalysis of glass by tephrochronology laboratories: results and recommendations, Quat. Int. 246 (2011) 19–47, http://dx.doi.org/10.1016/j.quaint.2011.08.022. [26] R.S. Boynton, Chemistry and technology of lime and limestone, WileyInterscience Ed, New York, 1996. [27] D. Miriello, D. Barca, A. Bloise, A. Ciarallo, G.M. Crisci, T. De Rose, C. Gattuso, F. Gazineo, M.F. La Russa, Characterisation of archaeological mortars from Pompeii (Campania, Italy) and identification of construction phases by compositional data analysis, J. Archaeol. Sci. 37 (2010) 2207–2223, http://dx.doi.org/10.1016/j.jas.2010.03.019. [28] A. Moropoulou, A. Cakmak, K.C. Labropoulos, R. Van Grieken, K. Torfs, Accelerated microstructural evolution of a calcium-silicate-hydrate (C-SH) phase in pozzolanic pastes using fine siliceous sources: comparison with historic pozzolanic mortars, Cem. Concr. Res. 34 (2004) 1–6, http://dx.doi.org/10.1016/S0008-8846(03)00187-X. [29] A. Moropoulou, A. Bakolas, S. Anagnostopoulou, Composite materials in ancient structures, Cem. Concr. Compos. 27 (2005) 295–300, http://dx.doi.org/10.1016/j.cemconcomp.2004.02.018. [30] E. Cantisani, A. Cecchi, I. Chiaverini, F. Fratini, C.M. Del Fa, E. Pecchioni, S. Rescici, The binder of the Roman concreteof the ponte di augusto at narni (Italy), Period. Di Mineral. 71 (2002) 113–123. [31] C. Genestar, C. Pons, A. Más, Analytical characterisation of ancient mortars from the archaeological Roman city of Pollentia (Balearic Islands, Spain), Anal. Chim. Acta. 557 (2006) 373–379, http://dx.doi.org/10.1016/j.aca.2005.10.058. [32] A.H. Munsell, Color soil charts, 1994. [33] A. Moropoulou, A. Bakolas, P. Michailidis, M. Chronopoulos, C. Spanos, Traditional technologies in Crete providing mortars with effective mechanical properties, Struct. Stud. Hist. Build. Tradit. Technol. 1 (1995) 151–161. [34] L. Barba, J. Blancas, L.R. Manzanilla, A. Ortiz, D. Barca, G.M. Crisci, D. Miriello, A. Pecci, Provenance of the limestone used in Teotihuacan (Mexico): a methodological approach, Archaeometry 51 (2009) 525–545. [35] M. de Gennaro, P. Cappelletti, A. Langella, A. Perrotta, C. Scarpati, Genesis of zeolites in the Neapolitan Yellow Tuff: Geological, volcanological and mineralogical evidence, Contrib. Mineral. Petrol. 139 (2000) 17–35, http://dx.doi.org/10.1007/s004100050571. [36] M. Jackson, D. Deocampo, F. Marra, B. Scheetz, Mid-Pleistocene pozzolanic volcanic ash in ancient Roman concretes, Geoarchaeology 25 (2010) 36–74, http://dx.doi.org/10.1002/gea.20295. [37] R. Fernández, B. Nebreda, R.V. De La Villa, R. García, M. Frías, Mineralogical and chemical evolution of hydrated phases in the pozzolanic reaction of calcined paper sludge, Cem. Concr. Compos. 32 (2010) 775–782, http://dx.doi.org/10.1016/j.cemconcomp.2010.08.003. [38] M.D. Jackson, S.R. Chae, S.R. Mulcahy, C. Meral, R. Taylor, P. Li, A.-H. Emwas, J. Moon, S. Yoon, G. Vola, H.-R. Wenk, P.J.M. Monteiro, Unlocking the secrets of Altobermorite in Roman seawater concrete, Am. Mineral. 98 (2012) 1669–1687, http://dx.doi.org/10.2138/am.2013.4484. [39] M. de Gennaro, A. Incoronato, G. Mastrolorenzo, M. Adabbo, G. Spina, Depositional mechanisms and alteration processes in different types of pyroclastic deposits from Campi Flegrei volcanic field (Southern Italy), J. Volcanol. Geotherm. Res. 91 (1999) 303–320, http://dx.doi.org/10.1016/S0377-0273(99)00040-2. [40] A. Colella, C. Di Benedetto, D. Calcaterra, P. Cappelletti, M. D’Amore, D. Di Martire, S.F. Graziano, L. Papa, M. de Gennaro, A. Langella, The Neapolitan Yellow Tuff: an outstanding example of heterogeneity, Constr. Build. Mater. 136 (2017) 361–373. [41] M.J. Le Bas, R.W. Le Maitre, A. Streckeisen, B. Zanettin, A chemical classification of volcanic rocks based on the total alkali-silica diagram, J. Petrol. 27 (1986) 745–750. [42] M. de Gennaro, C. Colella, M. Pansini, Hydrothermal conversion of trachytic glass into zeolite. II Reactions with high-salinity waters, Neues Jahrb. Far Mineral. 3 (1993) 97–110. [43] C. Rispoli, S.F. Graziano, A. De Bonis, P. Cappelletti, R. Esposito, P. Talamo, Piscina Mirabilis: characterization of geomaterials, 1st Int. Metrol. Conf. Archaeol. Ben evento (2015) 266–270. [44] D. Middendorf, B. KnOfel, Gypsum and lime mortars of historic German brick buildings: analytical results as well as requirements for adapted restoration material -, in: N.S.B. et al. (ed.), Conserv. Hist. Brick Struct. Case Stud. Reports Res. (1998). [45] C. Rispoli, S.F. Graziano, V. Guarino, A. De Bonis, C. Di Benedetto, R. Esposito, T. Budetta, V. Morra, P. Cappelletti, Characterization of ancient mortars: Preliminary results from VILLA DEL PEZZOLO, Sorrento Peninsula, Italy, in: IMEKO Int. Conf. Metrol. Archeol. Cult. Heritage. MetroArcheo 2016, 2016, pp. 151–156. [46] R. Zawawi, Artificial hydraulic lime mortar obtained by calcining limestone and siliceous waste materials, J. Adv. Appl. Ceram. 10 (2006) 175–178. [47] M.F. La Russa, S.A. Ruffolo, M. Ricca, N. Rovella, V. Comite, M.A. De Buergo, G.M. Crisci, D. Barca, Archaeometric approach for the study of mortars from the underwater archaeological site of Baia (Naples) Italy: preliminary results, Period. Di Mineral. 84 (2015) 553–567, http://dx.doi.org/10.2451/2015PM0031. [48] D. Miriello, A. Bloise, G.M. Crisci, R. De Luca, B. De Nigris, A. Martellone, M. Osanna, R. Pace, A. Pecci, N. Ruggieri, New compositional data on ancient mortars and plasters from Pompeii (Campania–Southern Italy): archaeometric

Please cite this article in press as: C. Rispoli, et al., The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy), Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.05.010

G Model CULHER-3605; No. of Pages 12 12

ARTICLE IN PRESS C. Rispoli et al. / Journal of Cultural Heritage xxx (2019) xxx–xxx

results and considerations about their time evolution, Mater. Charact. 146 (2018) 189–203, http://dx.doi.org/10.1016/j.matchar.2018.09.046. [49] R. De Luca, D. Miriello, A. Pecci, S. Domínguez-Bella, D. Bernal-Casasola, D. Cottica, A. Bloise, G.M. Crisci, Archaeometric Study of Mortars from the Garum Shop at Pompeii, Campania, Italy, Geoarchaeology 30 (2015) 330–351, http://dx.doi.org/10.1002/gea.21515. [50] F. Izzo, C. Grifa, C. Germinario, M. Mercurio, A. De Bonis, L. Tomay, A. Langella, Production technology of mortar-based building materials from the Arch of Trajan and the Roman Theatre in Benevento, Italy, Eur. Phys. J. Plus. 133 (2018) 1–12, http://dx.doi.org/10.1140/epjp/i2018-12229-1. [51] M.D. Jackson, S.R. Mulcahy, H. Chen, Y. Li, Q. Li, P. Cappelletti, H.R. Wenk, Phillipsite and Al-tobermorite mineral cements produced through low-temperature water-rock reactions in Roman marine concrete, Am. Mineral. 102 (2017) 1435–1450, http://dx.doi.org/10.2138/am-2017-5993CCBY.

[52] L. Sutter, Mineral characterization and cataloging of quarried aggregate sources, Volume I: final report submitted to the Michigan department of transportation, 2005. [53] G. Cultrone, E. Sebastián, M.O. Huertas, Durability of masonry systems: a laboratory study, Constr. Build. Mater. 21 (2007) 40–51, http://dx.doi.org/10.1016/j.conbuildmat.2005.07.008. [54] A. Arizzi, H. Viles, G. Cultrone, Experimental testing of the durability of limebased mortars used for rendering historic buildings, Constr. Build. Mater. 28 (2012) 807–818, http://dx.doi.org/10.1016/j.conbuildmat.2011.10.059. [55] E. Ontiveros-Ortega, O. Rodríguez-Gutiérrez, A.D. Navarro, Mineralogical and physical-chemical characterisation of Roman mortars used for monumental substructures on the Hill of San Antonio, in the Roman city of Italica (prov. Baetica, Santiponce, Seville, Spain), J. Archaeol. Sci. Rep. 7 (2016) 205–223, http://dx.doi.org/10.1016/j.jasrep.2016.03.043.

Please cite this article in press as: C. Rispoli, et al., The ancient pozzolanic mortars of the Thermal complex of Baia (Campi Flegrei, Italy), Journal of Cultural Heritage (2019), https://doi.org/10.1016/j.culher.2019.05.010