TY - JOUR
AU - Rizzo,, F
AB - Abstract In the last decade about 800 obsidian artifacts coming from various archaeological sites of Sicily have been analyzed using the BSC-XRF (beam stability controlled-x-ray fluorescence) and PIXE-alpha (particle induced x-ray emission, using low-energy alpha particles) portable spectrometers developed at the Landis laboratory of the LNS–INFN and IBAM–CNR in Catania (Italy). The portable BSC-XRF system allows the non-destructive analysis of Rb, Sr, Y, Zr and Nb trace concentrations, which are considered to be characteristic of the obsidian samples and consequently are indicative of the provenance quarries. Quantitative data on the above trace-element concentrations were deduced using a method that makes use of a multi-parameter linear regression. The portable PIXE-alpha spectrometer allows the quantitative determination of the matrix major elements, from Na to Zn. In this paper the updated versions of the instrumental devices and methods are presented together with a review of all the obtained data from various Sicilian sites. Results on compositional data for trace elements and major elements allowed us to identify Lipari and Pantelleria islands as the only two sources of the analyzed samples. Recent data about the Via Capuana settlement in Licodia Eubea are also presented and discussed for the first time. obsidian, provenance, PIXE, XRF 1. Introduction Obsidian is a well-known vitreous homogenous material that has been used in antiquity to produce different types of artifacts, including blades, bladelets and small flake tools. It is an intermediate to acid magmatic rock that comes from the rapid cooling of a magma rising to the surface. Its chemical composition is influenced by the magma differentiation processes (Wilson 1989). This is the main reason why some chemical elements, and in particular the trace elements Rb, Sr, Y, Zr and Nb, are considered to be characteristic of the magma evolution stages and, consequently, they can be considered as ‘fingerprints’ of the various volcanic sources. The study and the quantification of the chemical elements, and in particular of the above trace elements, in the obsidian artifacts allow researchers to establish the provenance of the raw material and, in some cases, the exchange routes (Cann and Renfrew 1964, Francaviglia 1999, Williams-Thorpe 1995, Crisci et al1994, Tykot 1996). In our laboratory, a program has been undertaken in the last few years to study the provenance of the obsidians recovered in different Sicilian archaeological sites. We have analyzed obsidians recovered in Ustica (Palermo) (Pappalardo 2013a)7 7 The authors thank Dr F Spatafora for the kind permission to refer on the preliminary results of a campaign of analysis of the obsidians recovered in the Villaggio dei Faraglioni. and (Tykot 1995), off the northern coast of Sicily, Milena (Caltanissetta) (La Rosa et al2006) in the center of Sicily, Rocchicella (Mineo) (Iovino et al2008, Pappalardo et al2008) in the central-east, Poggio dell’Aquila (Adrano) (Cultraro and Pappalardo 2010) and S. Marco (Paternò) (Maniscalco et al2008) on the western slope of Etna Mount, and Petraro (Siracusa) (Pappalardo 2013b)8 8 The authors thank Dr R Lanteri for the kind permission to refer on the Petraro, Vallone Macaudo, Case Vecchie obsidian analysis preliminary results. in the south-east (figure 1). Figure 1. Open in new tabDownload slide The Sicilian sites that have provided the obsidian artifacts in the LANDIS program. Grotta dell’Uzzo has been added. Lipari and Pantelleria islands are also shown. Licodia Eubea is situated in the south-east of Sicily. Figure 1. Open in new tabDownload slide The Sicilian sites that have provided the obsidian artifacts in the LANDIS program. Grotta dell’Uzzo has been added. Lipari and Pantelleria islands are also shown. Licodia Eubea is situated in the south-east of Sicily. Two non-destructive portable systems developed at the LANDIS laboratories, based on the XRF (x-ray fluorescence) and PIXE (particle induced x-ray emission) techniques, have been used in order to obtain a quantitative determination of the characterizing trace elements and, in some cases, of the light major ones. In the above-mentioned LANDIS study it has been possible to establish, for each site, the relative percentages of artifacts coming from Lipari and from Pantelleria; no obsidians coming from other Mediterranean sources have been found. In table 1 the above results are summarized, and it is evident how the use of non-destructive and portable instruments allowed the analysis of a large number of artifacts, even in situ. Table 1. The main results of the investigation on the provenance site for the obsidians analyzed in the LANDIS program. Site . Period . No of analyzed samples . Pantelleria % . Reference . Milena (CL): Serra del Palco, Middle 57 23% La Rosa et al Mandria Fontanazza-Monte Neolithic 2006 Grande Mezzebbi, Rocca dell’Aquila Copper Age Ustica (PA): Villaggio dei Faraglioni Middle Bronze Age 180a; 11 8%a 10% Pappalardo 2013a Tykot 1995 Mineo (CT): Rocchicella Palaeo-Mesolithic 53b 0% Iovino et al2008 Neolithic Copper Age 20b Pappalardo et al2008 Adrano (CT): Poggio dell’Aquila Copper Age 48 0% Cultraro and Pappalardo 2010 Paternò (CT): S Marco Late Neolithic 350 0% Maniscalco et al2008 Melilli (SR): Petraro, Vallone Maccaudo, Case vecchie Middle Neolithic 90a 0% Pappalardo 2013b Site . Period . No of analyzed samples . Pantelleria % . Reference . Milena (CL): Serra del Palco, Middle 57 23% La Rosa et al Mandria Fontanazza-Monte Neolithic 2006 Grande Mezzebbi, Rocca dell’Aquila Copper Age Ustica (PA): Villaggio dei Faraglioni Middle Bronze Age 180a; 11 8%a 10% Pappalardo 2013a Tykot 1995 Mineo (CT): Rocchicella Palaeo-Mesolithic 53b 0% Iovino et al2008 Neolithic Copper Age 20b Pappalardo et al2008 Adrano (CT): Poggio dell’Aquila Copper Age 48 0% Cultraro and Pappalardo 2010 Paternò (CT): S Marco Late Neolithic 350 0% Maniscalco et al2008 Melilli (SR): Petraro, Vallone Maccaudo, Case vecchie Middle Neolithic 90a 0% Pappalardo 2013b a Preliminary results. b Including five samples of Palagonite. Open in new tab Table 1. The main results of the investigation on the provenance site for the obsidians analyzed in the LANDIS program. Site . Period . No of analyzed samples . Pantelleria % . Reference . Milena (CL): Serra del Palco, Middle 57 23% La Rosa et al Mandria Fontanazza-Monte Neolithic 2006 Grande Mezzebbi, Rocca dell’Aquila Copper Age Ustica (PA): Villaggio dei Faraglioni Middle Bronze Age 180a; 11 8%a 10% Pappalardo 2013a Tykot 1995 Mineo (CT): Rocchicella Palaeo-Mesolithic 53b 0% Iovino et al2008 Neolithic Copper Age 20b Pappalardo et al2008 Adrano (CT): Poggio dell’Aquila Copper Age 48 0% Cultraro and Pappalardo 2010 Paternò (CT): S Marco Late Neolithic 350 0% Maniscalco et al2008 Melilli (SR): Petraro, Vallone Maccaudo, Case vecchie Middle Neolithic 90a 0% Pappalardo 2013b Site . Period . No of analyzed samples . Pantelleria % . Reference . Milena (CL): Serra del Palco, Middle 57 23% La Rosa et al Mandria Fontanazza-Monte Neolithic 2006 Grande Mezzebbi, Rocca dell’Aquila Copper Age Ustica (PA): Villaggio dei Faraglioni Middle Bronze Age 180a; 11 8%a 10% Pappalardo 2013a Tykot 1995 Mineo (CT): Rocchicella Palaeo-Mesolithic 53b 0% Iovino et al2008 Neolithic Copper Age 20b Pappalardo et al2008 Adrano (CT): Poggio dell’Aquila Copper Age 48 0% Cultraro and Pappalardo 2010 Paternò (CT): S Marco Late Neolithic 350 0% Maniscalco et al2008 Melilli (SR): Petraro, Vallone Maccaudo, Case vecchie Middle Neolithic 90a 0% Pappalardo 2013b a Preliminary results. b Including five samples of Palagonite. Open in new tab The most important results to emerge from the above investigation are: The raw material of obsidian recovered in the eastern part of Sicily is mainly from Lipari, while the one recovered in the western part shows a large contribution of Pantelleria. The percentage of artifacts from both sources and recovered in the archaeological Sicilian sites is strongly dependent on the source distance: the Grotta dell’Uzzo, which is located in the extreme western part of Sicily, has provided the larger percentage of Pantelleria obsidian, while the S. Marco, Petraro and Rocchicella sites, located in the eastern part, have provided only material from Lipari. Licodia (figure 1), located in the south-eastern part of Sicily, has provided a very large number (757) of obsidians allowing non-destructive analysis to be carried out on a wide range of samples. For this reason, the Licodia site seems to be an excellent case study in order to establish at which extent Pantelleria obsidian has reached the eastern part of the island. In the second section of the present work, a review is given of the BSC-XRF (beam stability controlled-x-ray fluorescence) and PIXE-alpha ((particle induced x-ray emission, using low-energy alpha particles)) experimental apparatus and of non-destructive analytical methods with emphasis on their recent developments in the LANDIS laboratory. The third section is devoted to the Licodia obsidian analysis, with particular attention to the raw material provenance and the archaeological background. Finally the conclusion is given, together with some hypotheses on the exchange routes of the Licodia obsidian. 2. The LANDIS laboratory of IBAM/CNR-LNS/INFN experimental apparatus employed for the non-destructive chemical analysis of obsidians Various geochemical techniques (such as conventional XRF, ICP–MS, SEM–EDS, conventional PIXE) have been employed to analyze obsidians. In general, they are non-portable and in some cases they are destructive. The LANDIS laboratory has developed two portable non-destructive systems based on the XRF and PIXE techniques which allow the discrimination between the Mediterranean sources (Iovino et al2008, Maniscalco et al2008, Pappalardo 2013a) giving one the opportunity to analyze the largest number of artifacts. However, mainly due to the limited number of analyzed elements and the limitation on beam intensity for safety reasons, the systems do not allow the discrimination of some sub sources. For the above reasons, the precision and the accuracy are, in general, lower compared to the laboratory fixed systems and/or destructive systems. 2.1. XRF technique The Rb, Sr, Y, Zr and Nb trace elements display some properties that enable them to be particularly suited to XRF analysis, currently performed with the usual low-current portable x-ray tubes working at 30–50 kV voltage and 0.2–0.5 mA current intensity, because: the high energy of their emitted x-ray K-lines (13.305 keV of Rb, 14.165 keV of Sr, 14.958 keV of Y, 15.775 keV of Zr, 16.615 keV of Nb) allows one to define an analytical depth of some hundred microns in the material, strongly reducing the contribution of eventual surface (few microns) effects (Liritzis 2006); the high ionization cross section (thousands of barns) induced by the incoming 20–30 keV x-ray beam favors the visibility of the corresponding x-ray lines in the energy spectrum; the intrinsic homogeneity of the obsidian due to its vitreous character gives a unique opportunity to perform quantitative analysis. The above described properties are well evident in the typical x-ray energy spectra taken on one obsidian artifact (figure 2) using a 45 kV voltage, 0.2 mA current intensity, Tungsten anode and EIS (Roma) x-ray tube. The characterizing trace elements lines are clearly visible together with the Fe and, less-pronounced, Ca ones. Figure 2. Open in new tabDownload slide Typical x-ray energy spectrum for an obsidian sample. Figure 2. Open in new tabDownload slide Typical x-ray energy spectrum for an obsidian sample. 2.1.1. The beam stability control and the BSC-XRF (beam stability controlled x-ray fluorescence) spectrometer To perform reliable quantitative XRF analysis, the stability of the outgoing beam intensity and energy is mandatory. The use of portable x-ray tubes, because of possible instabilities, requires continuous calibration procedures. To overcome this problem, in our laboratory a method to control the beam energy and intensity stability has been developed (Romano et al2005). The method is based on the observation that, for a given chemical element, a variation of the incident beam energy generates, due to the corresponding variation of the ionization cross section, a change in the intensity of the characteristic x-ray emitted lines. It is possible to identify two chemical elements for which the above intensity variation displays a different behavior with incident energy so that the ratio of the intensities can be used as an indicator of the energy variation. In the realized instrument (figures 3(a), (b)), two tin foils of Ag and Ba are positioned between the x-ray beam and the sample. Their K-lines are detected by an ancillary detector so that dedicated acquisition software (figure 4) can evaluate online the ratio of Ag and Ba intensity lines and their sum, allowing the control of the x-ray beam energy and intensity. Figure 3. Open in new tabDownload slide (a), (b) A recent version of the beam stability controlled (BSC-XRF) system apparatus. The double target (Ag and Ba) and the ancillary detector (Si PIN 20 mm2 active surface area) are visible. Also the main detector (Si Pin 20 mm2 surface active area) is shown. Figure 3. Open in new tabDownload slide (a), (b) A recent version of the beam stability controlled (BSC-XRF) system apparatus. The double target (Ag and Ba) and the ancillary detector (Si PIN 20 mm2 active surface area) are visible. Also the main detector (Si Pin 20 mm2 surface active area) is shown. Figure 4. Open in new tabDownload slide The front panel for the online evaluation of the ratio of Ag and Ba intensity lines and their sum, allowing the control of the x-ray beam energy and intensity. The oscillating curve is related to the variation of the x-ray beam energy. The tube voltage must be regulated so that the oscillating curve lies between the two pre-selected values. The straight line is related to the integrated x-ray beam intensity. A pre-set command assures the control of stability and the repeatability of the measurements. Figure 4. Open in new tabDownload slide The front panel for the online evaluation of the ratio of Ag and Ba intensity lines and their sum, allowing the control of the x-ray beam energy and intensity. The oscillating curve is related to the variation of the x-ray beam energy. The tube voltage must be regulated so that the oscillating curve lies between the two pre-selected values. The straight line is related to the integrated x-ray beam intensity. A pre-set command assures the control of stability and the repeatability of the measurements. 2.1.2. Multivariate treatment of the experimental results The obsidian artifacts were introduced to the spectrometer and x-ray energy spectra were obtained. The detector was a Si PIN AMPTEC 20 mm2 active surface. Rb, Sr, Y, Zr and Nb Kα-lines intensity peaks were extracted from the characteristic spectra through the deconvolution method of AXIL code and their relative values, that is the ratio of counts of each trace element peak to the total sum of counts of all trace element peaks, was determined. This procedure allows us to overcome the problems that could arise from the geometry of the samples and to compare the experimental data directly. These relative values were used to perform a multivariate cluster analysis (Hierarchical, Ward’s method, Euclidean distance), in all the studied cases. 2.1.3. The quantitative determination of the trace element concentration In order to individuate the provenance of the obsidian raw material, the concentrations of the five characterizing trace elements were deduced. From each group, we selected several samples possessing flat and smooth surfaces that were larger than the spot of the incident x-ray beam. The present employed method was first developed by Cirrincione and Pappalardo (1996) and was applied more recently to the analysis of Greek fine pottery and of powdered samples (Romano et al2006). It is based on the observation that: for a given matrix the ratio between the intensity and the concentration of each trace element is constant; the matrix composition in a silicate sample is uniquely established by the intensity of the most important groups of lines (in our case Ca and Fe). In other words silicate samples displaying the same intensity of the Ca and Fe x-ray lines are supposed to produce the same matrix effect with respect to the energetic x-rays emitted by the characteristic trace elements. 2.1.4. Calibration procedures and calculation of the concentrations BSC-XRF energy spectra on a set of geological reference standards have been used to extract the intensities NFe and NCa of the Fe and Ca x-ray Kα lines; the ratio k = c/N between concentration and intensity has been also deduced for the characterizing trace elements. A linear regression multi-parameter method has been employed to extract the parameters that give k in a function of NFe and NCa. Obviously, all the measurements have been carried out in the same experimental conditions (beam intensity, voltage, measurement time, geometry). The obtained calibration equations were (Romano et al2006): k Rb =0.058+3.19×10-5N Ca +1.43×10-6N Fe k Sr =0.059+1.99×10-5N Ca +7.04×10-7N Fe kY=0.037+2.11×10-5N Ca +4.61×10-7N Fe k Zr =0.033+9.69×10-6N Ca +5.87×10-7N Fe k Nb =0.029+1.05×10-6N Ca +5.83×10-7N Fe .1 The above equations have been used to calculate the concentrations in the unknown cases by inserting the values of NFe and NCa intensities net counts into equation (1) to deduce the k values; the concentration has been calculated as c = kN, N being the intensity of line of a given trace element. 2.2. The particle induced x-ray emission technique XRF has shown to be a reliable and simple method to establish the provenance of obsidian raw material by trace-element analysis. However, we have decided to extend the analysis to the light major elements by performing PIXE measurements, in a few typical cases showing a large and flat surface. In the Licodia case, we have used the portable PIXE-alpha system (Pappalardo et al2003) of LNS/INFN and CNR/IBAM. This system makes use, as a charged particle emitter, of an annular 210Po α-radioactive source. Recently (Pappalardo et al2013) the system has been upgraded by using a new detector which allows an x-ray energy resolution of 124 eV @ 5.9 KeV and a large area 210Po source (Romano et al2012) an efficiency tree times larger than the previous one has been reached with a beam spot diameter of 10 mm. Quantitative analysis has been carried out using the GUPIXWIN code (Maxwell et al1995, Hopman et al2002). The PIXE analytical method is suited for obsidians due to the homogeneity of the vitreous material. 3. The case of the Via Capuana settlement in Licodia Eubea 3.1. The archaeological background During 1992 and 1995, the Soprintendenza ai Beni Culturali ed Ambientali di Catania directed the excavation of a Late Neolithic village recovered in the urban centre of Licodia Eubea, within the Iblaean Mountains area, in south-eastern Sicily (Palio 2012). The settlement has an extension of about 1000 sqm and, on the basis of the stratigraphic evidence, two phases of life have been identified. The existence of huts is indicated only by scant remains of floors. Two big structures seem to have had a central role in the life of the village—a large quadrangular area with a clay floor in the central portion of the settlement and a long curved furrow dug on the rock surface, perhaps the foundation trench of a large apsidal hut (figure 5). Figure 5. Open in new tabDownload slide Plan of the Via Capuana Neolithic settlement (Licodia Eubea, excavations 1992). Figure 5. Open in new tabDownload slide Plan of the Via Capuana Neolithic settlement (Licodia Eubea, excavations 1992). The site has provided more than 8000 pieces of flint, including tools and debitage, and a very large group of obsidians (757 artifacts), whose typological and technological analyses are presently in progress (Bracchitta 2011). The flint and obsidian items were widespread throughout the area, though a higher density was registered south of the large flat clay ground in the central area of the excavation. It is worth emphasizing that the 757 obsidians represent not only the larger ‘exotic’ commodity found in Licodia but, more precisely, the highest quantity of obsidian artifacts ever gathered from Late Neolithic layers in Sicily. They include formalized blades and bladelets (84%), whereas the debitage is very scant (14%). It is also important to remark upon the presence of eight obsidian cores (corresponding to the remaining 1% of the obsidian assemblage) and five flakes with small cortex remains, which suggests that the raw material was partially imported in rough nodules and that the reduction process in some cases had taken place within the village itself. The obsidians’ typology is mostly represented by unretouched blades and bladelets realized with uniform typometric criteria. The striking platform or the distal end were often removed from the blades. In some cases, the striking platforms are flat or more commonly pointed, as they were produced from unidirectional cores respectively by using indirect percussion and pressure flaking techniques. Multidirectional cores are attested as well, from which mainly unretouched flakes were detached (figure 6). Figure 6. Open in new tabDownload slide (1) Flake; (2) unidirectional core (refitted); (3, 5, 6) retouched blades; (4, 7, 8, 9, 10) unretouched blades; (11, 12, 13) retouched flakes (drawings by D Bracchitta). Figure 6. Open in new tabDownload slide (1) Flake; (2) unidirectional core (refitted); (3, 5, 6) retouched blades; (4, 7, 8, 9, 10) unretouched blades; (11, 12, 13) retouched flakes (drawings by D Bracchitta). Determining the provenance of the different obsidian artifacts gathered from the site of Via Capuana is an issue of great interest in order to throw light on the contacts that may have existed, even though indirectly, between the Neolithic community of Licodia and the overseas obsidian sources located many kilometers away (Palio 2012, Bracchitta 2011). 691 randomly chosen items, from the 757 obsidian artifacts, have been studied. 3.2. Results 3.2.1. BSC-XRF measurements From the BSC-XRF data and from the cluster analysis two main groups (group 1, containing 682 cases, and group 2, containing 9 cases) were obtained, suggesting that the volcanic glasses come from only two different sources. Group 2 was subdivided in two sub-groups (group 2b, containing four cases and group 2c, containing five cases) (figure 7). Table 2 and figures 8(a), (b) show the average values of the obtained concentrations for the two groups and their comparison to the data from Lipari and Pantelleria sources. Figure 7. Open in new tabDownload slide (a), (b), (c) Typical x-ray energy spectra of three samples: one sample belonging to group 1 and two samples belonging to groups 2b and 2c. Figure 7. Open in new tabDownload slide (a), (b), (c) Typical x-ray energy spectra of three samples: one sample belonging to group 1 and two samples belonging to groups 2b and 2c. Figure 8. Open in new tabDownload slide (a), (b) A comparison between the calculated data and those in the literature, from Lipari (a) (Francaviglia 1999, Williams-Thorpe 1995) and Pantelleria (b) (Francaviglia 1999, Williams-Thorpe 1995) sources. The data on subsources Balata dei Turchi, Sciuvechi and Lago di Venere are also shown. Figure 8. Open in new tabDownload slide (a), (b) A comparison between the calculated data and those in the literature, from Lipari (a) (Francaviglia 1999, Williams-Thorpe 1995) and Pantelleria (b) (Francaviglia 1999, Williams-Thorpe 1995) sources. The data on subsources Balata dei Turchi, Sciuvechi and Lago di Venere are also shown. Table 2. Average values of the trace elements’ concentration (ppm) obtained in the present work for the different obsidian groups compared to the known data (Francaviglia 1999) from Lipari and Pantelleria. Errors on the results (arising mainly from the calibration procedure, the repeatability of the data (Romano et al2006) and counting statistics) are about 10% for Rb, Y, Zr and Nb. In the present case due to counting statistics, Sr is at the limit of detection. . Rb . Sr . Y . Zr . Nb . Group 1 303 13  52  204  48 Group 2b 181  6 183 2065 318 Group 2c 119  8  96 1065 156 Lipari, (Papesca) 319 14  59  210  27 Pantelleria, (Sciuvechi) 188  3 176 1815 354 Pantelleria, (Lago di Venere) 113  1  97 1015 211 . Rb . Sr . Y . Zr . Nb . Group 1 303 13  52  204  48 Group 2b 181  6 183 2065 318 Group 2c 119  8  96 1065 156 Lipari, (Papesca) 319 14  59  210  27 Pantelleria, (Sciuvechi) 188  3 176 1815 354 Pantelleria, (Lago di Venere) 113  1  97 1015 211 Open in new tab Table 2. Average values of the trace elements’ concentration (ppm) obtained in the present work for the different obsidian groups compared to the known data (Francaviglia 1999) from Lipari and Pantelleria. Errors on the results (arising mainly from the calibration procedure, the repeatability of the data (Romano et al2006) and counting statistics) are about 10% for Rb, Y, Zr and Nb. In the present case due to counting statistics, Sr is at the limit of detection. . Rb . Sr . Y . Zr . Nb . Group 1 303 13  52  204  48 Group 2b 181  6 183 2065 318 Group 2c 119  8  96 1065 156 Lipari, (Papesca) 319 14  59  210  27 Pantelleria, (Sciuvechi) 188  3 176 1815 354 Pantelleria, (Lago di Venere) 113  1  97 1015 211 . Rb . Sr . Y . Zr . Nb . Group 1 303 13  52  204  48 Group 2b 181  6 183 2065 318 Group 2c 119  8  96 1065 156 Lipari, (Papesca) 319 14  59  210  27 Pantelleria, (Sciuvechi) 188  3 176 1815 354 Pantelleria, (Lago di Venere) 113  1  97 1015 211 Open in new tab From the data of table 2 and figure 8 we deduce that the Licodia obsidian raw material is manly of Lipari provenance. Only nine samples can be attributed to Pantelleria sources. In particular, the raw material of five of them is consistent with that indicated by Francaviglia (1988, 1999) as of Lago di Venere origin. The recently published data of D’Amora et al (2012) on Pantelleria subsources concern ICP-MS analytical results on some known and new Pantelleria obsidian subsurces: Balata dei Turchi, Cala della Polacca and Salto della Vecchia are chemically very similar and cannot be distinguished by our analytical systems, so that in the graphs of figure 8(b) their data coincide, within the instrumental uncertainties, with the Balata dei Turchi (Francaviglia 1988, 1999) data. Lago di Venere and Faraglioni di Dietro, which are similar but both chemically different from the first ones, are non-distinguishable from the Lago di Venere (Francaviglia 1999) present data. 3.2.2. PIXE-alpha measurements The samples were placed in contact with the head of the instrument. The acquired spectra were analyzed with the use of the GUPIX software in the oxide option. The H constant was determined by the use of the SCO-1 reference standard. The measurement time was 3600 s. The major light element composition also confirms the existence of the two main groups already evidenced by the trace element XRF data. It must be noted that the two groups differ mainly by the Fe content, which is much higher for the group that corresponds to the Pantelleria sources (figure 9). Figure 9. Open in new tabDownload slide (a), (b) The obsidian light elements’ concentration, as obtained from the PIXE-alpha technique, compared to the literature data from Lipari and Pantelleria. The provenance suggested by the XRF analysis is confirmed. In particular, the high presence of Fe is a further indication of the peralcaline composition of the Pantelleria obsidians. Figure 9. Open in new tabDownload slide (a), (b) The obsidian light elements’ concentration, as obtained from the PIXE-alpha technique, compared to the literature data from Lipari and Pantelleria. The provenance suggested by the XRF analysis is confirmed. In particular, the high presence of Fe is a further indication of the peralcaline composition of the Pantelleria obsidians. 4. Conclusion The data of table 1, together with the new Licodia data, are presented in the map of figure 10, in which the green bars indicate the Pantelleria provenance and the gray ones the Lipari one. No other provenances have been found. Figure 10. Open in new tabDownload slide Provenance and distribution of obsidian artifacts in Sicily. Figure 10. Open in new tabDownload slide Provenance and distribution of obsidian artifacts in Sicily. Concerning the obsidians coming from Licodia, 682 samples come from Lipari sources (our spectroscopic systems are not sensitive to the chemistry of the various Lipari sources as they have been distinguished, for example, in Tykot et al2006) and nine samples come from Pantelleria sources. A careful analysis of these latter items has shown that some of the Pantelleria obsidians (five samples) can be associated with that of Lago di Venere. The other 4 samples are associated with different Pantelleria sources that cannot be distinguished by our spectroscopic non-destructive methods, as they have been distinguished, for example, in Francaviglia (1998), Tykot (1995) and, more recently, by D’Amora et al (2012). Yet, during the present investigation, one small pebble of Palagonite has been discovered among the obsidian artifacts, and it is characterized by a very typical XRF x-ray spectrum (Iovino et al2008). 4.1. Archaeological considerations The most unexpected and meaningful result of the present study is the ascertained provenance of nine obsidian artifacts from Pantelleria lava flows. Although these specimens represent a very exiguous part of the whole assemblage, they overturn the previous released pattern of obsidian exchange network due to the geographical position of the site (Nicoletti 1997). In the following paragraphs we have tried to outline some hypotheses on obsidian internal exchange along eastern Sicily. 4.2. Lipari provenance As concerns the obsidian from Lipari, a reliable hypothesis could view the Via Capuana site as the tail end of a wide land obsidian network that connected this site to the northern coast of Sicily. It is more probable that the products coming from Lipari arrived here, then were transported through the Peloritani mountain passes, as can be inferred from the Eneolithic obsidian assemblage from the site of Castiglione La Marca, close to the Alcantara river (Nicoletti 1997), and then through the Simeto valley where several Late Neolithic villages reveal clues about obsidian-core processing (Maniscalco 2003). The large quantity of obsidian artifacts (757) found in Licodia and the presence of debitage can be explained by considering the village as a site devoted to knapping activity, situated above the richest flint exploitation district in Sicily. The most part of the obsidian assemblage is not retouched (98%), and this might indicate that the obsidian items found within the great flint assemblage of Via Capuana may be the result of intense interactions and exchanges relating to the movement of the Iblean flint in Sicily. 4.3. Pantelleria provenance The presence of Pantelleria obsidian in Licodia might be explained by the assumption of relations with the villages of western and central Sicily, as it is the nearest landing place for sailors who came from the isle of Pantelleria. In this part of Sicily, in fact, the percentages of Pantelleria obsidian are noticeably higher (Tykot 1996) and, as a general behavior, it is evident that the Pantelleria obsidian presence decreases (in percentage with respect to the total number of samples) depending on the source distance: it is largely attested in the western area of Sicily (Grotta dell’Uzzo (40%), Milena (23%), Ustica (10%) and at a minimum rate in Licodia (1,2%). It is remarkable that, with the only exception of Licodia, the sites located East of the Salso river have not provided evidence of Pantelleria provenance. Acknowledgment Thanks are due to Professor G Pappalardo for enlightening discussions and comments. References Bracchitta D . , 2011 Overseas lithic exchanges between Aeolian Islands and Malta from an inland perspective: preliminary data from a Late Neolithic site in Licodia Eubea, Catania—Sicily SOMA: Proc. 15th Symp. on Mediterranean Archaeology 3–5 March University of Catania, Italy at press Cann J R , Renfrew C . , 1964 The characterisation of obsidian and its application to the Mediterranean region , Proc. Prehistoric Soc. , vol. 30 (pg. 111 - 133 ) Google Scholar Crossref Search ADS WorldCat Cirrincione R , Pappalardo L . , 1996 Trace element composition as fingerprint for non destructive and in situ analysis of ancient materials Proc. 13th Int. Congress of Prehistoric and Protohistoric Sciences Forlì 8–14 Sept. (pg. 305 - 311 ) Crisci G M , Ricq-De Bouard M , Lanzafame U , De Francesco A M . , 1994 Nouvelle Méthode d’analyse et provenance de l’ensemble des obsidiennes néolithiques du Midi de la France , Gallia Préhistoire , vol. 36 (pg. 299 - 327 ) 10.3406/galip.1994.2129 Google Scholar Crossref Search ADS WorldCat Crossref Cultraro M , Pappalardo G . , 2010 L’industria su ossidiana dall’insediamento preistorico sul Poggio dell’Aquila presso Adrano (Catania): analisi non distruttive e aspetti morfo-tipometrici , Dialogo dei Saperi—Metodologie Integrate per i Beni Culturali (IBAM-CNR) Tomo I D’Andria F , Malfitana D , Masini N , Scardozzi G . Naples Edizioni Scientifiche Italiane (pg. 393 - 400 ) Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC D’Amora A , Tufano E , Trifuoggi M , Tusa S . , 2012 Le potenzialità dell’applicazione del metodo ICP-MS allo studio dei reperti paletnologici. Il caso dell’ossidiana di Pantelleria Atti della 41th Riunione Scientifica dell’Istituto Italiano di Preistoria e Protostoria San Cipirello (Pa), Firenze (pg. 1197 - 1403 ) Francaviglia V . , 1988 Ancient obsidian sources on Pantelleria (Italy) , J. Archaeol. Sci. , vol. 15 (pg. 101 - 122 ) 10.1016/0305-4403(88)90001-5 Google Scholar Crossref Search ADS WorldCat Crossref Francaviglia V . , 1999 The search for a reliable discrimination between Mediterranean obsidians Proc. Il vetro in Italia meridionale e insulare 5–7 March. 1998 Napoli Sogliani F , Piccioli (Napoli) C . (pg. 315 - 360 ) Hopman T L , Nejedly Z , Maxwell J A , Campbell J L . , 2002 Extension of GUPIX to 2H, 3He and 4He excitation , Nucl. Instrum. Methods Phys. Res. B , vol. 189 (pg. 138 - 142 ) 10.1016/S0168-583X(01)01019-9 Google Scholar Crossref Search ADS WorldCat Crossref Iovino M R , Maniscalco L , Pappalardo G , Pappalardo L , Puglisi D , Rizzo F , Romano F P . , 2008 Archaeological volcanic glass from the site of Rocchicella (Sicily, Italy) , Archaeometry , vol. 50 (pg. 474 - 494 ) 10.1111/j.1475-4754.2007.00352.x Google Scholar Crossref Search ADS WorldCat Crossref La Rosa V , Palio O , Pappalardo G , Pappalardo L , Romano F P . , 2006 Analisi delle provenienze di ossidiane del territorio di Milena (CL) Atti della 39th Riunione scientifica dell’Istituto Italiano di Preistoria e Protostoria: Materie Prime e scambi nella preistoria italiana Firenze 25–27 Nov. 2004 (pg. 499 - 508 ) Liritzis I . , 2006 SIMS-SS, a new obsidian hydration dating method: analysis and theoretical principles , Archaeometry , vol. 48 (pg. 533 - 547 ) 10.1111/j.1475-4754.2006.00271.x Google Scholar Crossref Search ADS WorldCat Crossref Maniscalco L . , 2003 Il neolitico nella media e bassa valle del Simeto Atti della 25th Riunione Scientifica dell’Istituto Italiano di Preistoria e Protostoria—Le comunità della Preistoria italiana—Studi e Ricerche sul Neolitico e sull’Età dei Metalli 2–7 Giugno 2000 Castello di Lipari, Chiesa S. Caterina, Firenze (pg. 963 - 968 ) Maniscalco L , Massimino A , Motta L , Pappalardo L , Pappalardo G , Romano F P , Rizzo F . , 2008 L’ossidiana di S. Marco (Paternò) nel contesto dei ritrovamenti siciliani Scienza e Beni Culturali: Atti del V Congresso Nazionale di Archeometria 26–29 Febbraio 2008 Siracusa Syracuse Morrone (pg. 483 - 491 ) Maxwell J A , Teesdale W J , Campbell J L . , 1995 The Guelph PIXE software package II , Nucl. Instrum. Methods Phys. Res. B , vol. 95 (pg. 407 - 421 ) 10.1016/0168-583X(94)00540-0 Google Scholar Crossref Search ADS WorldCat Crossref Nicoletti F . , 1997 Il commercio preistorico dell’ossidiana nel mediterraneo e il ruolo di Lipari e Pantelleria nel più antico sistema di interscambio , Prima Sicilia Tusa (Palermo) S . (pg. 259 - 269 ) Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Palio O . , 2012 L’insediamento tardo-neolitico di via Capuana a Licodia Eubea (CT). Dai Ciclopi agli Ecisti. Società e territorio nella Sicilia preistorica e protostorica Atti della 41th Riunione Scientifica dell’Istituto Italiano di Preistoria e Protostoria San Cipirello (PA), Firenze (pg. 535 - 542 ) Pappalardo G . , 2013a private communication Pappalardo G . , 2013b private communication Pappalardo G , Massimino A , Pappalardo L , Romano F P , Rizzo F . , 2008 Recenti risultati sul materiale litico vetroso , Il Santuario dei Palici Un centro di culto nella valle del Margi Maniscalco L . Palermo Assessorato dei Beni Culturali, ambientali e della pubblica Istruzione Regione Siciliana (pg. 401 - 408 ) Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Pappalardo L , Alberti R , Calì C , Garraffo S , Litrico P , Pappalardo G , Rizzo F , Romano F P . , 2013 The new PIXE-alpha spectrometer for the analysis of Roman nummi surfaces , X-ray Spectrom. , vol. 42 (pg. 33 - 37 ) 10.1002/xrs.2429 Google Scholar Crossref Search ADS WorldCat Crossref Pappalardo L , Romano F P , Garraffo S , De Sanoit J , Marchetta C , Pappalardo G . , 2003 The improved LNS PIXE-ALPHA portable system: archaeometric applications , Archaeometry , vol. 45 (pg. 333 - 339 ) 10.1111/1475-4754.00111 Google Scholar Crossref Search ADS WorldCat Crossref Romano F P , Calvi G , Furia E , Garraffo S , Marchetta C , Pappalardo G , Pappalardo L , Rizzo F , Rovelli A . , 2005 A new portable XRF spectrometer with a beam stability control to be used in the cultural heritage field , X-ray Spectrom. , vol. 34 (pg. 135 - 139 ) 10.1002/xrs.785 Google Scholar Crossref Search ADS WorldCat Crossref Romano F P , Pappalardo G , Pappalardo L , Garraffo S , Gigli R , Pautasso A . , 2006 Quantitative non-destructive determination of trace-elements in archaeological pottery using a portable beam-stability controlled XRF spectrometer , X-ray Spectrom. , vol. 35 (pg. 1 - 7 ) 10.1002/xrs.880 Google Scholar Crossref Search ADS WorldCat Crossref Romano F P , Pappalardo L , Calvi G , Costa E , Marchetta C , Pappalardo G , Rizzo F , Russo S . , 2012 A new version of a portable polonium source for the non-destructive PIXE (particle induced X-ray emission) analysis in the cultural heritage field , Microchem. J. , vol. 101 (pg. 95 - 98 ) 10.1016/j.microc.2011.11.007 Google Scholar Crossref Search ADS WorldCat Crossref Tykot R . , 1995 Appendix I: obsidian provenance , Ustica I. The results of the excavations of the Regione Siciliana, Soprintendenza ai beni culturali ed ambientali provincia di Palermo in collaboration with Brown University in 1990 and 1991 Archaeologia Transatlantica vol 14 Holloway R R , Lukesh S S . Providence, RI Center for World Archaeology and Art, Brown University, Département d’archéologie et d’histoire de l’art, Collège Erasme (pg. 87 - 90 ) Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Tykot R H . , 1996 Obsidian procurement and distribution in the central and western Mediterranean , J. Mediterr. Archaeol. , vol. 9 (pg. 39 - 82 ) Google Scholar Crossref Search ADS WorldCat Tykot R H , Iovino M R , Martinelli M C , Beyer L . , 2006 Ossidiana da Lipari: le fonti, la distribuzione, la tipologia e le tracce d’usura Atti della 39th Riunione Scientifica dell’Istituto Italiano di Preistoria e Protostoria: Materie prime e scambi nella preistoria siciliana 25–27 Novembre Firenze (pg. 592 - 597 ) Williams-Thorpe O . , 1995 Obsidian in the Mediterranean and the Near East: a provenancing success story , Archaeometry , vol. 37 (pg. 217 - 248 ) 10.1111/j.1475-4754.1995.tb00740.x Google Scholar Crossref Search ADS WorldCat Crossref Wilson B M . , 1989 , Igneous Petrogenesis London Unwin Hypman 10.1007/978-1-4020-6788-4 Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Crossref © 2013 Sinopec Geophysical Research Institute
TI - Obsidian provenance determination using the beam stability controlled BSC-XRF and the PIXE-alpha portable spectrometers of the LANDIS laboratory: the case of the Via Capuana settlement in Licodia Eubea (Sicily)
JF - Journal of Geophysics and Engineering
DO - 10.1088/1742-2132/10/6/064004
DA - 2013-12-01
UR - https://www.deepdyve.com/lp/oxford-university-press/obsidian-provenance-determination-using-the-beam-stability-controlled-e8llMBU2bb
VL - 10
IS - 6
DP - DeepDyve
ER -