A survey on innovative dating methods in archaeometry with focus on fossil bones

ARTICLE IN PRESS Trends in Analytical Chemistry ■■ (2015) ■■–■■

Contents lists available at ScienceDirect

Trends in Analytical Chemistry j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / t r a c

A survey on innovative dating methods in archaeometry with focus on fossil bones Mauro Tomassetti *, Federico Marini, Remo Bucci, Luigi Campanella Department of Chemistry, University of Rome “La Sapienza”, Rome, Italy

A R T I C L E

I N F O

Keywords: Dating archaeological finds Innovative dating methods Fossil bones Chemometrics Thermal analysis

A B S T R A C T

Recently proposed innovative methods for dating archaeological finds reported in the literature are reviewed, together with the researches carried out in this field by our team, using several instrumental techniques (sensor, biosensor, electrochemical, thermal-analytical, and so on). In this framework, particular attention was then focused on examining the main currently adopted methods for bones dating. Lastly, the possibility of using thermogravimetry coupled to chemometrics for differentiating, quickly and inexpensively, finds of very ancient human fossil bones (several thousand years BC), from less ancient ones (few hundreds of years before and after Christ) is illustrated, together with a thorough discussion of the criteria (i.e., the collagen-carbonates content ratio, or the different types of carbonate) this approach is based on. © 2015 Elsevier B.V. All rights reserved.

Contents 1. 2. 3.

Innovative dating of archaeological finds, using sensors, biosensors, other electrochemical methods, thermal analysis, biochemical clocks and beyond . Principal methods reported in the literature for ancient bones dating and characterization .................................................................................................... Dating fossil bones through thermal analysis and chemometrics ........................................................................................................................................................ Acknowledgements ............................................................................................................................................................................................................................................... References ................................................................................................................................................................................................................................................................

1. Innovative dating of archaeological finds, using sensors, biosensors, other electrochemical methods, thermal analysis, biochemical clocks and beyond As it is well known, the fossil dating is essential to develop a chronometric scale applicable in sketching a geological history of the stratigraphic classification of rocks and for dating geological events. In practice, the fossil dating can offer a time-scale for relative age determinations and for world-wide correlations of rocks. This kind of relative dating method can be considered a rather old standard approach. However, since the 60s of the last century, absolute dating methods, particularly radiometric, have also been developed. The majority of them is, in fact, based on the decay of radioactive isotopes, either the so-called “primordial” ones, such as 238U, 230Th, or 40 K, or of so-called “continuous creation” radioisotopes, such as 14C. Despite effectively suffering from numerous drawbacks, also these methods have now become classical approaches, like the rest of other methods now very popular, such as the Fission Track,

* Corresponding author. Tel.: +39 06 49913722; Fax: +39 06 49913725. E-mail address: [email protected] (M. Tomassetti).

1 4 6 8 8

Thermoluminescence, or Optically Stimulated Luminescence, Electron Spin Resonance, Dendrochronology, Obsidian dating, Amino acids racemization and Archaeomagnetism. For these techniques, which have become well established, the applicability range and the accuracy are rather well characterized (even if in some cases the different sources are only partially agreeing) and are summarized in Table 1. Of course, in the meantime, several researchers have attempted to develop, and then to propose, other approaches, different from those mentioned, certainly still much less known, and not fully investigated in terms of their applicability characteristics, and received with greater or less attention, for several reasons. For example our research team has been developing and testing instrumental chemical method and chemometric techniques in order to meet different needs, prompted by archaeologists and paleontologists, concerning dating, differentiation, classification, or the characterization of different types of archaeological, paleontological, or cultural heritage findings. These objectives have been addressed with different approaches: for instance, making use of biosensing, electrochemical or enzymatic methods. Among them, it is worth mentioning a biosensor method [31,32] for the dating of wood or paper finds, based on the construction of an amperometric biosensor, assembled using a Clark-type gaseous

http://dx.doi.org/10.1016/j.trac.2015.11.017 0165-9936/© 2015 Elsevier B.V. All rights reserved.

Please cite this article in press as: Mauro Tomassetti, Federico Marini, Remo Bucci, Luigi Campanella, A survey on innovative dating methods in archaeometry with focus on fossil bones, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.11.017

ARTICLE IN PRESS M. Tomassetti et al./Trends in Analytical Chemistry ■■ (2015) ■■–■■

2

Table 1 Summary of the characteristics of the classical dating approaches in archaeometry Method

Chronological interval (years)

Accuracy (Relative error %)

References

Pb-U series

102-5 × 103* 5 × 103-5 × 104 5 × 104-3 × 105 3 × 105-106 103-104* 104-5 × 104 5 × 104-105 105-107 3 × 102-104* 104-5 × 104 5 × 104-3 × 105 3 × 105-107 102-5 × 102 5 × 102-2 × 103 2 × 103-5 × 104 102-5 × 103 5 × 103-3 × 104 3 × 104-2 × 106 2 × 106-107* 102-4 × 105 4 × 105-106* 102-6 × 103 102-3 × 103 3 × 103-8 × 103* 3 × 104-9 × 104* 9 × 104-107 102-104 104-4 × 105 4 × 105-4 × 106*

25–75% 8–25% 2–8% 8–25% >75% 25–75% 8–25% 2–8% >75% 25–75% 8–25% 2–8% >8% 2–8% <2% >75% 25–75% 8–25% 25–75% 8–25% 25–75% <2% 8–25% 25–75% 25–75% 8–25% 25–75% 8–25% 25–75%

[1–6]

Potassium-Argon

Fission track

Radiocarbon

Thermoluminescence/ Electron spin resonance

Obsidian hydration Dendrochronology Archaeomagnetism Palaeomagnetism Amino acid racemization

[1,3,4,7]

[1,3,4,8,9]

[1,3,4,10–21]

[1,3,4,22,23]

Fig. 1. Definition of the environmental persistence through the photosensor response. The environmental persistence (Penv) is defined through the slope (sTiO2) and the delay time (t) as: Penv = t/sTiO2.

[1,3,4,24,25] [1,3,4,26] [1,3,4,27] [1,3,4,28] [1,3,4,29]

* Indicates extended range under special experimental conditions [30].

diffusion amperometric electrode and the enzyme glucose oxidase, immobilized by carbodiimide (N- (3-propyl-dimethylamino)-N’ethylcarbodiimide) on a standard disk, cut from the specimen to be dated. The principle of the method is the following: since it has been shown [33] that the more ancient a cellulosic sample is, the greater the number of carboxy groups contained in it, and, therefore, the more the molecules of a given enzyme (e.g. glucose oxidase) which can be covalently immobilized on it, with a simple method, such as that of carbodiimide. Accordingly, if one prepares a disk of cellulosic specimen (wood, cloth, paper) with a standardized size and weight, the older the sample is, the higher the amount of glucose oxidase covalently immobilized on it will be. This diskette is then fixed on the head of an amperometric electrode (Clark-type), which is then immersed in a buffer solution, to which a fixed quantity of the enzyme substrate (i.e., glucose) is added. As a consequence of the enzymatic reaction, which consumes the oxygen in solution, a decrease of the current intensity through the electrode will be observed; such decrease is proportional to the amount of immobilized enzyme, and, hence, to the number of carboxylic groups present on the find, therefore to the sample’s antiquity. For the same types of objects, but also for plant tissues, a different enzymatic method was also experienced by us [34], introducing some innovations to the one originally developed by S. Kouznetsov [33], and based on the enzyme SAMT (S-adenosylmethionine-transmethylase), capable of catalyzing a reaction of transferring methyl groups, present on a ancient cellulosic find, to a specific acceptor molecule. In this enzymatic reaction, adenosine, which can be easily determined with different instrumental methods [33–35], is generated and the antiquity of cellulosic analyzed find evaluated on the basis of the quantity of adenosine checked. In detail, the cellulosic material, in the presence of the enzyme SAMT, S-adenosylmethionine (SAM) and Tetrahydropholic acid (THFA), is converted to demethylated cellulosic material, adenosine and omocysteine.

Another dating method, proposed by our group, consisted in the development of a potentiometric photosensor, based on the catalyst titanium dioxide (TiO 2 ) and a platinum electrode [36]. A potentiometric curve is obtained by means of this sensor, during the photodegradation of a cellulosic artifact, by irradiation with UV light (λ = 350 nm) in a reaction cell. In particular, due to the chemical species that are formed in solution during the catalytic degradation of the sample under investigation, a variation of the electrode potential, which occurs according to a characteristic pattern (see Fig. 1), which allows to measure a parameter called “environmental persistence”, is produced; the latter is put in relation with the aging state and, then, with the antiquity of a find, especially cellulosic [36–38]. On the other hand, the use of cyclic voltammetry, has recently allowed not only to determine the amount of lignin [39], but also to make a short archaeometric curve of the same lignin artificially aged [39]. These cited methods have dealt especially with the dating of cellulosic finds. It is clear that several different electrochemical methods have also been proposed by other authors in the literature. From a general point of view, all classical Faradic electrochemical methods are very sensitive techniques for identifying and determining a lot of electroactive specie present in the samples. They are able to carry out speciation studies, providing also a complete description of the state of oxidation of the ionic species contained in the finds. Also non-Faradic electrochemical methods, for instance conductometric techniques, have been extensively used, in several occasion, e.g. for monitoring the content of salts removed during water immersion treatments of ancient ceramic remains and archaeological potteries. In this way the electrochemical techniques, being able to characterize in detail the materials concerned, are certainly able to guide and assist the dating of archaeological finds, even when this is done with methods more specific for this purpose. In addition to these general electroanalytical methods, recently it has been evidenced that the chemical evaluation of the extent of corrosion in archaeological artifacts can provide insight into the duration of the corrosion process, so that the latter process, by using several simplifying assumptions, can be used to estimate the antiquity of a metal archaeological find. Reich et al. [40] firstly used measurements on the Meissner fraction in the superconducting state to evaluate the mass of the uncorroded metal sample, contained in the find, in order to estimate the age of the archaeological lead artifacts. In brief, lead metal, at room temperature exhibits diamagnetic susceptibility of the same order of magnitude

Please cite this article in press as: Mauro Tomassetti, Federico Marini, Remo Bucci, Luigi Campanella, A survey on innovative dating methods in archaeometry with focus on fossil bones, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.11.017

ARTICLE IN PRESS M. Tomassetti et al./Trends in Analytical Chemistry ■■ (2015) ■■–■■

as its salts and oxides. However, when cooled below 7.2 K, it enters into the superconducting state, in which it exhibits the ideal diamagnetic susceptibility. The value of this susceptibility is much higher than that of lead oxide and salts, which are no superconductors. The determination of value of the volume magnetic susceptibility of lead in the Meissner state (i.e. a state in which there is a total expulsion of the magnetic field from the volume of the lead sample) permits to calculate the mass of lead metal effectively existing in the sample, as the molar susceptibility of lead oxides and salts down the cryogenic temperatures is negligible in comparison to the diamagnetic signal of lead. So that, knowing the total mass of the sample, the mass of the corrosion products can be determined. Lastly, dating metal objects is so possible, assuming that aging occurred, under uniform conditions, by estimating the growth of the mass per surface unit of the corrosion layer. Based on this principle, electrochemical methodologies were proposed by Scholz et al. [41], which exploit the capabilities of the voltammetry of microparticles for the analysis of metal finds [41,42] and for dating these metal objects, by estimating the extent of the corrosion process starting from the recording of characteristic electrochemical signatures of the compounds forming the primary and secondary patinas in lead and copper bronze artifacts [43–45]. Using this method A. Doménech-Carbó et al. (2011) [43] dated archaeological lead artifacts, measuring the time variation of voltammetric signatures specifically attributed to PbO2 and PbO species; or for dating archaeological copper-based artifacts, determining the tenorite/cuprite ratio, by exploiting the production of well-differentiated voltammetric responses by such compounds [45]. The same approach was used by A. Doménech-Carbó et al. for dating porcine blood-based binding media used in Taiwanese architectural polychromes [46]. In this research the voltammetry of microparticles was used to date haemoglobin – based archaeological materials, by recording the voltammetric response of the material to be dated, attached to paraffin–impregnate graphite electrodes in contact with aqueous acetate buffer. Signals attributable to the FeIII/ FeII iron couple, and their catalytic enhancement in the presence of H2O2, can be correlated with the time of aging of the sample. Lastly the voltammetry of microparticles methodology was used, by A. Doménech-Carbó et al. [44], to date archaeological lead pieces, from the funds of different Spanish museums, based on the time – dependent formation of different layers of lead oxides, whose relative amount can be estimated from polarization curves and electrochemical impedance spectroscopy (EIS) and additional data were obtained using square wave voltammetry. Another electrochemical dating method, reported by F. Scholz et al. [47], describes a technique which enables the detection of electronic point defects, i.e. trapped electrons and holes in non conducting materials. The method, which, from the standpoint of the assumptions it is based on, is not different, for instance, from the well known dating technique that measures the thermoluminescence, is based upon the phenomenon that oxygen evolution and reduction is catalyzed on the surface of oxide phases which contain electronic defects. The authors demonstrated that specific signals occur when electronic defects are formed as a result of radiation damage. This signal was measured by the authors of the published work on solid samples, with abrasive stripping voltammetry, or using modified carbon paste electrodes. In the first case the solid sample was immobilized on the surface of a solid electrode, whereas in the second case, the powdered sample was mixed with graphite and oil to make an electrode. Cyclic voltammetry was used. Different ceramic finds were attempted to dating using these techniques, obtaining satisfactory results. However, cellulosic finds have been studied by our group, not only using electrochemical methods, but also by thermal analysis. Indeed, we tried to obtain archaeometric curves, using thermogravimetric (TG) and derivative (DTG) techniques, coupled

3

to kinetic and chemometric methods [48]: in these cases, the parameters considered were, in a first stage and for wooden objects [49], the percent cellulose/lignin ratio, or the onset temperature; in a later stage, studying, besides wood, also other kind of findings, especially paper, more “elaborate” methods have been developed, which rely on the calculation of the activation energy (E) of the cellulose (contained in the find) degradation process, using the kinetic methods of Arrhenius, or Satava [48,50], or the method of Wyden and Widmann [48,51,52]. Also other important authors used thermal analytical techniques to dating archaeological finds. Wiedemann and Bayer [53] investigated the swelling behaviour of papyri in relation to age, by subjecting a number of papyri to TMA (Thermal Machanical Analysis) and showed how the expansion of the swelling papyri curves decreases, increasing the age of papyri. H.G. Wiedemann [54] studied in depth papyri and tree wood using TG, DTG, DTA (Differential Thermal Analysis), and DSC (Differential Scanning Calorimetry) and observed that the cellulose/lignin ratio of papyri of different ages changes with the time. The same is true for the trees; for the latter, however, also the effect of climatic fluctuations is reflected. According to this author, the position and the height of the DTG peak of papyri are very sensitive to its age, but they may also be affected by the processing of the papyrus and the state of its preservation. In addition the DSC curves of ancient Egyptian papyrus show that the lignin peak decreases slightly with increasing age of the sample. Besides, in a section of very old tree (giant sequoia) the shape and height of the lignin DSC peak could be assigned unambiguously to the age of the annual growth ring; while both the cellulose and the lignin DTG peaks are only slightly shifted to lower temperature with the age, the heat of combustion increases in the same order and lignin shows less pronounced changes with age. F. Preußer [55] has shown that it is possible to date paintings, created within the last 100 years, from the height of DTA peaks of natural substances used to make them, such as rosin, beeswax and especially albumen. Odlyha and Burmester [56] studied by DTA some binding media of paintings, such as oil, egg yolk and their mixture with pigments (e.g. linseed oil/egg yolk/ZnO), or (linseed oil/lead white/ some addition of proteinaceous material). These authors concluded that, on the basis of the DTA and DTG peaks of minuscule fractions of the paintings it is possible to provide indications of the age and characterize the binding media of studied paintings, prepared in the studied period (1915–1941). Lastly Odlyha et al. [57–59] has shown that it is possible to obtain accurate archaeometric trends, with representations by means of histograms, both of suitable specimens artificially aged, and of veritable paintings, conserved in various international museums, using data obtained by DSC, but especially by means of DMA (Dynamic Machanical Analysis). A dating method that has aroused considerable interest in recent years and has given rise to the publication of a good number of works, is the one generally called “Biochemical clocks” [1]. This method, also called molecular clocks, uses the molecular changes associated with mutations to estimate the time passed since two species diverged genetically. Usually nucleotide sequences of DNA, or aminoacid sequences of proteins, were used for such genetically calculation. Some observations of Zukerkandl and Pauling [60] may be considered the basis of this method but M. Kimura [61], carrying out studies on haemoglobin molecules among different groups of animals, suggested that, during the evolutionary history of mammals, amino acids substitution has taken place roughly, at the rate of one amino-acid change in 107 years, for a chain consisting of 140 amino acids. In practice, this author assumed that there is a constant rate of mutation that occur in a definite population of individuals. F.J. Ayala [62] claimed that molecular evolution is dependent on the fickle process of natural selection, but it is a time – dependent process, so that accumulation of empirical data often

Please cite this article in press as: Mauro Tomassetti, Federico Marini, Remo Bucci, Luigi Campanella, A survey on innovative dating methods in archaeometry with focus on fossil bones, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.11.017

ARTICLE IN PRESS 4

M. Tomassetti et al./Trends in Analytical Chemistry ■■ (2015) ■■–■■

yields an approximate clock, as a consequence of the expected convergence of large numbers. S.Y.W. Ho et al. [63] declared that studies based on population – level and pedigree data, have produced remarkably high estimates of mutation rate, which strongly contrast with substitution rates inferred in phylogenetic studies. The same authors evidenced the relationship between the age of the calibration and the rate of change can be described by a vertically translated exponential decay curve, which may used for correcting molecular data estimates. Consequently the results of several previous studies need, to be revaluated. S. Horwath [64] conducted a massive study in this field to establish if DNA methylation levels can be used to accurately predict age across human tissues and cell types, or whether the resulting age prediction is a biologically meaningful measure. Finally, he proposed that the DNA methylation age measures the cumulative effect of an epigenetic maintenance system. This novel epigenetic clock can be used to address a host of questions in developmental biology, cancer and aging. A.J. Drummond [65] affirmed that the unrooted model of phylogeny and the strict molecular clock model, are two extremes of a continuum, and despite their dominance in phylogenic inference, it is evident that both are biologically unrealistic, lastly that the real evolutionary process lies between these two extremes. Fortunately intermediate models, employing relaxed molecular clocks, have been described, which open the gate to a new field of relaxed phylogenetics. Lastly the author introduced a new approach to performing relaxed phylogenetic analysis, which is phylogenetically more accurate and precise than the traditional unrooted model, while adding the ability to infer a timescale to evolution. Concluding this brief review dedicated to the use of molecular clocks it can be observed how the use of this method is constrained by a series of factors [62]. Accordingly the use of molecular clocks requires taking into account a series of conditioning factors [63–66]. When the studies of the authors of the present paper had as their object stone artifacts, the role of chemometric data processing was more fundamental, while the instrumental methods of analysis were of a different kind: Plasma Emission (ICP), infrared spectroscopy (IR), X-ray diffraction (XRD) and thermal analysis (TG and DTA). In these cases, the problem was generally to differentiate and characterize mainly marbles [67], or terracotta finds [68,69] and to identify their provenance, that sometimes allow also the individuation of the age; in the latter case, thermomechanical (TMA) and porosimetric analyses were also used [69]. All these researches have definitely evidenced how the thermal analytical data can be particularly suitable to be processed by chemometric techniques. Therefore, other types of artifacts, such as pigments [70], mortars from Roman or Renaissance frescoes [71] and, more recently, fossil human bones [72,73], have been studied coupling these two methods (see Section 3 of the present paper). The study of fossil bones, in particular, is of great importance both from an archaeological and a paleontological standpoint; therefore, a great host of methods, of various types, have been developed in the last years. 2. Principal methods reported in the literature for ancient bones dating and characterization A first not chemical but important dating method is called biostratigraphy; it uses “index-fossils” plus the “law of superposition”. This method can yield pretty good dates – but only relative. Independently, considerable effort has been expanded, starting of about the 1950, with reference to the chemical changes which occur in fossil bones, during diagenesis and several works on the subject of bone fossilization have been issued [74,75]. During prolonged diagenesis, the relative percent content of three primary analyzable inorganic components of bones (calcium, phosphate and carbonate)

experiences significant alteration. Usually, there is an increase in all the per cent by weight of those substances, due to the progressive loss of the organic matter. In particular, a sharp increase in carbonate was observed by several authors [76–78], owing that carbonate was accumulated by precipitation from soil solution during the fossilization process. However the carbonate accumulation from different sites varies within wide limits [78]. Of course what is true for carbonate is likewise true also for calcium and phosphate during the diagenesis: some bones show heavy accumulation some little change and some even a loss, according the burial’s soil. Therefore the chemical history of every site differs from that of every other one and, consequently, the bones buried must be different with respect to their inorganic constituents; this issue represents the great obstacle to the chemical method of dating. Nevertheless, several dating methods have been proposed. A first method relies on the comparison of the unknown bone specimen with a sample of another bone, the date of which can be securely established. If the composition of the two bones is the same, within a reasonable analytical error, they can be considered contemporaneous [79]. On the other hand, upon the discovery made in the nineteenth century that animal bones, lying in the ground, accumulate fluorine and incorporates this element into the crystal structure, giving fluoroapatite, Oakley [80] utilized the fluorine content of human bones as a criterion of age. Unfortunately, since the rate of a accumulation of this element is a function not only of time, but also of the fluorine concentration in the soil in which the bone is buried, it is not possible to build a standard archaeometric scale to be used in every site [81]. In addition, in 1956 it was found that, in the mineral phase of bones saturated with fluorine, the amount of fluorine present is not a simple function of hydroxyl ion replacement, but there is some other process going on simultaneously which could not be only defined by crystallographic methods [82]. Moreover, there is a considerable variation in the amount of fixed fluoride, when moving from the external surface progressively towards the interior of the bone [83]. Lastly, the advent of radioisotopes led to the possibility that new micro methods for the determination of traces of fluoride could be developed. A very appropriate method would be the standard isotope dilution; taking into account however that the half-life of 18 F is only 112 min, probably mass spectrometry can offer a simplified tool for quantitative analysis of the small amounts of fluoride contained in bones; however, at present, this possibility is still rarely adopted [81]. Another element, which has been treated in a similar way of fluorine, is uranium. In this context, an Interesting study was reported by some authors [2], which proved a clear tendency toward accumulation of this element. However, also the uranium concentration varies from soil to soil, so that the method may suffer from the same drawbacks above cited for fluorine. Another chemical technique for dating bones is the so called “bone-nitrogen dating technique”. This technique is based on the consideration that bones buried in soil lose organic components (causing, in turn, a decrease of the amount of nitrogen), while, at the same time, inorganic components such as fluorine and uranium are gained. Since bones buried at the same time, in the same deposit will lose nitrogen and gain fluorine and uranium at the same rate, analytical chemical methods able to measure these components can be uses as a relative dating technique to determine if bones found in the same matrix were indeed deposited together. Although bone– nitrogen dating technique can not produce an exact age, as the rate of nitrogen loss and, for instance, fluorine gain can differ with local environmental condition, nonetheless, when used in conjunction with other bone dating methods, this approach can allows accurately dating a collection of bones, if the information about the age of a few specimen in the collection is available [84]. Lastly, only a short mention can be made concerning a variety of methods that yield actual calendrical dates for fossils. More

Please cite this article in press as: Mauro Tomassetti, Federico Marini, Remo Bucci, Luigi Campanella, A survey on innovative dating methods in archaeometry with focus on fossil bones, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.11.017

ARTICLE IN PRESS M. Tomassetti et al./Trends in Analytical Chemistry ■■ (2015) ■■–■■

actually, they date the soil layer the bones are in, or associated materials, and not the fossil of interest. A number of them are radiometric methods, that make use of the decay of radioactive isotopes. Uranium is a common trace element, and various isotopes decay in various patterns. The best for dating is the ratio thorium/ uranium (230Th / 234U). Perhaps the most common radiometric dating techniques is potassium-argon dating. The method has been helpful in dating formation associated with remains of fossil hominides and lower Paleolithic tools, for instance the remains of zinjanthropus to approximately 1,750,000 years ago [84]. Other absolute dating methods of radiometric kind, but which may not be suitable for bone dating, are thermoluminescence and electrospin resonance. Differently to the inorganic components, it was generally conceded that the organic component of bones do not interchange with soil constituents, but tend to diminish with the antiquity. Therefore, there is the possibility that the total organic fraction, or certain specific organic substances, may decompose independently of the burial soil composition. Some observations reported by Barber [77] and Pin [85], who studied the total nitrogen content of bones of different age, seemed to confirm this idea, and Thunberg [86] examined a series of recent and prehistoric bones and analyzed their citric acid content. On the basis of his results some discrepancies have been recognized, so that the same author stated that citric acid is not a valid criterion for determining bone age and, later on, serious divergences have been shown even in the attempting to correlate the aminoacids content with the antiquity of bone find. There has been clearly established a differential removal of organic matter, including fat, citric acid, protein, carbon, nitrogen and probably also other organic materials. Therefore a simple, unequivocal dating system can not be based upon the analyses of the organic constituents of fossil bones. However in relatively more recent years two methods received much attention and became very popular. The first was named “amino acid racemization” method. Its foundation is the fact that amino acids exist in two forms (optical isomers), known as (L) and (D). All biological organisms, living on the Earth planet, use only the (L) form, but, after death, aminoacids begin to be partly converted to the (D) form until they reach equilibrium at a 1:1 ratio. This process is well known and named “racemization”. Accordingly, by experimentally measuring the D/L ratio of a chosen amino acid, using chromatographic techniques, i.e. HPLC or GC and chiral columns, when the rate of conversion is known, racemization provides a clock that can be used to determine the time of death of whatever biological organism. Of course, the amino acid should be carefully selected, as the racemization rate is different for substance. A typical example of accurate dating of fossil bones using the racemization method was published by J.L. Bada [29]. In this case the racemization of isoleucine was used to estimate the age of fossil bones. Results suggest that the isoleucine racemization reaction may provide an important tool for dating of fossil bones, including bones which are too old to be datable by radiocarbon. However, the major limitation of this approach is that the rate of racemization reaction depends also on temperature, so that some estimate of the temperature history of the fossil bone must be available for a correct application of the method. The second relatively recent method was the radiocarbon method; in short the unstable isotope 14C is formed when cosmic rays hit 14 N in the atmosphere. Organisms take it up during life, along with the common 12C. Once an animal dies, the 14C is not replenished and the ratio 14C/12C drops. Knowing the half-life, by comparing the atmospheric ratio to the specimen ratio, it is possible to calculate the age of the specimen when it died. By now this method is so important that the origin of bone dating goes right to the earliest publications on the 14C method, i.e. the well known “radiocarbon dating” [10,11].

5

The earliest dates were derived from total bone carbon, but the reliability of these early measurements was immediately questioned owing to the frequently scarce correlation with results of stratigraphic method. Therefore the attention was focused on determining whether fractions of bone could be identified that would provide reliable dates. The principal preparation methods have tended to focus the organic fraction, principally collagen. Collagens are a family of proteins and the major type (type I collagen) is the primary structural protein in bone. Most radiocarbon dates on bone, teeth and tusk, are derived from radiocarbon content measurement of collagen in the organic phase. However depending upon the burial environment, hydrolytic reactions can fragment the polypeptides, degrading collagen molecules to other biomolecules, so that covalently linking contaminants into the collagen matrix may occur; therefore several methods have been developed for collagen extraction from bones and teeth. Berger et al. [12] first proposed a collagen isolation method based upon the suspension of bone powder in diluted strong HCl acid. The insoluble washed fraction was obtained without substantially reducing collagen yield, but is often contaminated with humic or fulvic acids. Longin [13] proposed a simple modification to this early collagen purification protocol, introducing, after collagen extraction in HCl solution, a gelatinization step, that rendered the collagen into soluble gelatine (heating an aqueous suspension of collagen at temperature greater than 65°C, upon coaling, the solution forms a gel). The gelatinization changes the game, so that any remaining insoluble contaminants can be removed from the collagen solution by either centrifugation or filtration. Successively a modified Longin method was introduced, which provides an extraction in NaOH solution, after the first extraction in HCl solution to eliminate the soilderived organic matter contaminants that are insoluble in acidic solution. The latter extraction in NaOH was followed by another acid wash to ensure that any absorbed atmospheric carbon dioxide is removed before the gelatinization step. This modified Longin method is commonly referred to as an “ABA plus gelatinization” method. Finally the gelatinized collagen can be isolated from neutral pH solutions by freeze-drying. This dried collagen solid is easy to handle or store. Brown et al. [14] further amended the Longin or modified Longin method, introducing a subsequent ultrafiltration using 30 kDa molecular weight cut off membranes. The retained fraction, containing only high molecular weight collagen molecules, is lastly dated. A collagen total amino acid dating has been proposed by Ho, Marcus and Berger [15], which proposed that the purified collagen by Longin method, was hydrolized in 6N HCl at 100°C for 24 h. The amino acid–containing residue was then purified by ion– exchange chromatography. The total amino acids eluted were subsequently combusted for radiocarbon measurement. An alternative to the total collagen dating, was the targeting of specific amino acids, particularly hydroxyproline, uncommon in most proteins, but contained into collagen. Its rarity made it a good marker for collagen. Several investigators focused on developing methods for hydroxyproline isolation method, for instance Stafford et al. [16], Gillespie et al. [17], Mc Cullagh et al. [18] and Marom et al. [19]. Finally Nelson [20] formulated an interesting method, extensively utilized by Tisnérat-Laborde et al. [21]. In the first step crude collagen extracts are treated with ninhydrin in a slightly acid buffered solution. The ninhydrin–treated collagen polypeptides are recovered and then hydrolyzed to their constituent amino acids in hot 6N HCl. When this preparation is treated a second time with ninhydrin, the resulting carbon dioxide gas is recovered for radiocarbon dating. Besides the more classical dating method above described, other different methods are recently proposed. For instance a technique based X-ray diffraction [87]. Shortly experimental results showed that the older the bone the higher is its cristallinity and the sharper is it X-ray diffraction pattern (XRD). To this end a cristallinity index

Please cite this article in press as: Mauro Tomassetti, Federico Marini, Remo Bucci, Luigi Campanella, A survey on innovative dating methods in archaeometry with focus on fossil bones, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.11.017

ARTICLE IN PRESS 6

M. Tomassetti et al./Trends in Analytical Chemistry ■■ (2015) ■■–■■

was experimentally checked that might be used as a measure of the degree of fossilization and could be used also as a method for relative dating. An other interesting method to absolute dating of bone finds concerns the possibility of absolute dating of bone finds by measuring the intensity of the emitted autofluorescence through the use of a confocal microscope [88]. Of course these innovative methods are of a preliminary nature and further development of these approaches as a dating methods is necessary. 3. Dating fossil bones through thermal analysis and chemometrics As evidenced in the previous paragraph, the “classical” methods for dating fossil samples, and in particular bones, may suffer from different drawbacks and therefore the possibility of using also other approaches, which can lead to a more simple, cheap, fast, but objective identification of the age of the findings or, at least, to a clear differentiation on the basis of the samples’ antiquity represents a promising alternative. In this context, while only few and restrict applications report the use of some special spectroscopic techniques, such as vibrational spectroscopy of fossil [89], a term that encompasses Fourier Transform, Infrared (FTIR) and Raman spectroscopy; or the Near Infrared Hyperspectral Chemical Imaging (NIR-HCI) [90], to rapidly and non destructively evaluate the relative degree of collagen preservation in bones from archaeological finds, a major role is played by the possibility of using thermal analytical methods coupled by chemometric processing of the collected data. Thermal analytical curves, resulting from thermogravimetry (TG) or derivative thermogravimetry (DTG), represent a wealth of precious information, which may result useful for the differentiation of bone samples on the basis of their age. Indeed, when looking the thermogravimetric profiles of bone samples, three distinct windows may be identified, corresponding to thermal transition involving the loss of water, and the decomposition of collagen and carbonates, respectively. In particular, as shown in Fig. 2, where typical TG/ DTG curves recorded on fossil bone samples are reported, the first step (A), which is linked to moisture loss, is associated to a DTG peak at about 100°C. On the other hand, collagen decomposition (B) may either occur as a single transition (DTG peak at about 335°C) or appear as two substeps, corresponding to DTG peaks at about 330–350 and 450–500°C, respectively. At higher temperatures, between about 600 and 850°C, several different TG-DTG steps are all linked to carbonate decomposition. Based on this information, a first approach to using thermal analytical data for assessing the different antiquity of bone samples

Fig. 2. TG (a) and DTG (b) curves of several fossil bone samples from Middle Nile necropolises, showing the main thermal analytical transitions.

Fig. 3. Example of bidimensional Szoor’s plot based on the thermogravimetric data measured on fossil bone samples of different age from Middle Nile necropolises. The abscissa (A + B) is calculated as the sum of the mass loss due to moisture (A) and to collagen decomposition (B) while the fossilization coefficient Fk, reported in the ordinate is defined as the quotient between (A + B) and the mass loss associated to carbonate decomposition (C). More ancient samples (Mesolithic) are plotted as red squares while the more recent ones (meroitic and Christian) are represented as black circles).

was proposed by Szoor [91], who observed that, with increasing bone sample age, the percentage of carbonate contained in the samples increased while, at the same time, the percentage of collagen decreased. Such observation, lead him to propose a graphical approach based on the definition of two parameters, calculated on the basis of the mass losses of the three main thermogramivetric steps: a coefficient (A + B) calculated as the sum of the mass loss due to moisture (A) and to collagen decomposition (B), and a fossilization coefficient Fk, defined as the quotient between (A + B) and the mass loss associated to all the transitions involved in carbonate decomposition (C). Accordingly, Szoor evidenced that, by reporting Fk vs (A + B), samples of different ages give rise to distinct clusters in the bidimensional plot (see Fig. 3). More recently, Tomassetti et al. [73] have questioned the need of including the term A in both the coefficients proposed by Szoor, by pointing out that, due to storage issues before analysis, the moisture content could be an unreliable index, and have shown that by plotting B/C vs B a clearer separation among samples of different age could be obtained. However, one of the limitation of such an approach is that it provides an oversimplistic representation of the thermal transitions occurring in the bone samples, as both the collagen and the carbonate decompositions may be the results of thermogravimetric steps involving different forms of the those substances. Indeed, as far as collagen is concerned, the possible presence of either one (step B1, at about 350°C) or two (the former plus a second one near 450–500°C, indicated as step B2) DTG peaks was already evidenced by L.F. Lozano et al. [92], who performed a thorough study of the thermal stability of human modern and ancient bones. In particular, the presence of the second DTG peak at higher temperature was ascribed by these authors to the fact that part of the collagen contained in the bone may present a different configuration of chains, involving a higher number of crosslinks or secondary bonds. Such an hypothesis was further supported by Tomassetti et al. [72] on the basis of the calculations of the activation energies (Ea) of the two thermogravimetric transitions associated to collagen decomposition: while the Ea of the substep at lower temperatures is of the order of 10 kJ/mol, the one of the second decomposition process is about 17 kJ/mol, which is perfectly in line with the presence of a higher degree of collagen crosslinking (higher stability). In this framework, the possible presence of either one or both transition was explained by Sakae et al. [93] by suggesting that the progress of the fossilization process induce a corresponding disruption of the more

Please cite this article in press as: Mauro Tomassetti, Federico Marini, Remo Bucci, Luigi Campanella, A survey on innovative dating methods in archaeometry with focus on fossil bones, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.11.017

ARTICLE IN PRESS M. Tomassetti et al./Trends in Analytical Chemistry ■■ (2015) ■■–■■

cross-linked collagen, so that the more ancient samples present a single DTG peak at lower temperature, as also confirmed by Tomassetti et al. [72]. On the other hand, above approximately 600°C, smaller decomposition steps may be observed, all ascribable to the thermal decomposition of the carbonate. In particular, the most relevant contribution is given by the so-called “re-formation” or “re-precipitation” carbonate (step C1) [85,93], which decomposes between 600 and 750 °C in one or more substeps (the transition temperatures depending on the size of crystals) and whose relative amount increases with the age of fossil bones (Lozano et al. [92], E. Bonucci and G. Graziani [94] and above all by G. Szoor [91]). A further decomposition step is observed between about 750°C and 850°C (step C2, giving rise to a small and broad DTG peak), which may be ascribed to the decomposition of the carbonate ion originally present in the apatite lattice [Haas and Banewicz] [95]. Accordingly, following the ideas of Haas and Banewicz [95], Tomassetti et al. [73] have proposed different ways of combining the mass loss due to reformation carbonate and that due to the breakdown of the carbonate originally present in the apatite lattice into parameters which could be graphically represented in a bidimensional plot similar to the one proposed by Szoor [91]: the combination tested allowed a moderately good separation of samples of different antiquity. Based on the above consideration, it was proposed [72] that a more accurate differentiation among bone samples coming from different ages could be achieved by building mathematical (chemometric) models which took into account a higher number of variables summarizing the characteristics of the thermal analytical profiles. In particular, whenever possible, it was proposed to build a data matrix comprising the transition temperatures and the mass losses of all the main decomposition steps and substeps identifiable in the TG/DTG curves, and to process such matrix using a suitable chemometric tool. In this respect, since the number of bone findings available is almost always too low to allow building a reliable predictive (classification) model to be used for dating, an unsupervised approach based on principal component analysis is often recommended. Indeed, principal component analysis allows a parsimonious representation of multivariate data, by projecting the samples onto a low dimensional subspace, at the same time capturing as much as possible of the variance originally present in the data. As a result, one obtains two sets of plots [72] which allow, respectively, graphically inspecting the relations among samples and intepreting the observed differences in terms of the measured chemical and physicochemical variables. In particular, the representation of the projection of the samples onto the space spanned by the principal components is called the scores plot, and it allows to evidence differences and similarities among the objects, and even possible trends. Indeed, it may be possible that samples with similar characteristics appear as more or less clearly identifiable groups or clusters in the scores plot, allowing a partial or complete differentiation. As an example, the scores plot obtained by principal component analysis of the main thermal analytical data extracted by TG/DTG profiling of several bone samples of different ages from necropolises in the Middle Nile (Sudan) is reported in Fig. 4. It is possible to observe in the Figure how the analyzed samples are grouped in two distinct clusters, one likely corresponding to Mesolitic and the other corresponding to more recent (Meroitic and Christian) bones, evidencing how the interplay between thermal analysis and chemometrics may provide a straightforward comparative dating of the findings. The other result of principal component analysis is the so-called loadings plot, i.e. a graph reporting the relative contribution of each of the experimentally measured variables to the definition of the projection directions (the principal components). Accordingly, inspection of the loadings plot provides the interpretation of the differences observed in the scores plot on the basis of the experimentally measured

7

Fig. 4. PCA analysis of the main TG-DTG data (% mass losses of the main thermal transitions, DTG peak temperatures and residual) of fossil bones samples of different age from Middle Nile necropolises after autoscaling. Projection of the analyzed samples onto the space spanned by the first and second principal components (scores plot). More ancient samples (Mesolithic) are plotted as red squares while the more recent ones (meroitic and Christian) are represented as black circles).

variables. The loadings plot reported in Fig. 5, which corresponds to the same principal component model used for building the scores plot in Fig. 4, suggests that the major contribution to the separation of the two groups of samples is related to the mass loss of collagen and carbonates and to the residue at 1000°C. Although this approach has proved to be successful in different studies, the extraction of the required features (peak temperatures and mass losses) from the thermal analytical curves may not always be easy and accurate, especially when multiple partially overlapping transitions are involved. Therefore, another possibility [72] is to carry out the chemometric analyses directly on the whole TG or DTG curves, thus avoiding the need of feature extraction. In general, such an approach may provide a clearer differentiation among the samples (better separated clusters) but, since the whole profile is used to describe the data, the interpretation of the loadings may not be as straightforward, as in the case when the analysis is carried out on the features [72,73]. In particular, this last approach allowed to address almost unequivocably some concerns expressed by Mkukuma et al. [96] on the use of the information of collagen decomposition for the dating of fossil bones. Indeed, these authors suggested that the

Fig. 5. PCA analysis of the main TG-DTG data (% mass losses of the main thermal transitions, DTG peak temperatures and residual) of fossil bones samples of different age from Middle Nile necropolises after autoscaling. Projection of the experimental variables onto the space spanned by the first two principal components (loadings plot).

Please cite this article in press as: Mauro Tomassetti, Federico Marini, Remo Bucci, Luigi Campanella, A survey on innovative dating methods in archaeometry with focus on fossil bones, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.11.017

ARTICLE IN PRESS M. Tomassetti et al./Trends in Analytical Chemistry ■■ (2015) ■■–■■

8

observed differences in the collagen content may be ascribed not only to the different antiquity of the findings but also on the different anatomical part of the skeleton the bone samples were withdrawn from. On the other hand, in a recent paper, Tomassetti et al. [97] have shown that, by coupling TG/DTG and chemometrics it was possible to evidence that the different anatomical parts of the skeletons seem, in fact, not to contribute significantly to the differentiation of the samples, especially in the case of the older ones, which are more mineralized. Among the more recent ones, a further differentiation can perhaps in part be sought at most in the interindividual variability, between one skeleton and another, rather than in the various anatomical parts of a same skeleton [97]. As a last remark, it should be stressed that, although very promising, the described approach can not be used yet to obtain an exact dating of the analyzed samples nor to build any archaeometric scale, as this would require a further validation step, involving the comparison of the outcomes with those obtainable using a chronological dating method, which, so far, couldn’t be carried out. Acknowledgements The research was funded by the University of Rome “La Sapienza” (Progetto di Ricerca 2014; grant nr. C26A14YTEC). References [1] A. Doménech-Carbó, Dating: an analytical task, ChemText 1 (2015) 5 1-5.29. [2] E.B. Jaffe, A.M. Sherwood, Physical and Chemical Comparison of Modern and Fossil Tooth and Bone Material, U.S. Geological Survey, Oak Ridge, TN, 1951. [3] K.L. Pierce, Dating methods, in: Geophysics Research Forum (U.S.) Geophysics Study Committee, “Active tectonics, The National Academy Press, Washington, DC, 1969, pp. 195–214. [4] M.J. Aitken, Science-Based Dating in Archaeology, Routledge, Abingdon, UK, 2013. [5] M. Ivanovich, R.S. Harmon (Editors), Uranium Series Disequilibrium: Application to Environmental Problems, Clarendon Press, Oxford, 1982. [6] A. Kaufman, W.S. Broecker, T.L. Ku, D.L. Thurber, The status of U-series methods on mollusk dating, Geochim. Cosmochim. Acta 35 (1971) 1155–1183. [7] G.B. Dalrymple, M.L. Lanphere, J. Gilluly, A.O. Woodford, Potassium-Argon Dating: Principles, Techniques, and Applications to Geochronology, W. H. Freeman, San Francisco, CA, 1969. [8] G.A. Wagner, P. Van den Haute, Fission-Track Dating, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1992. [9] C.W. Naeser, B. Bryant, M.D. Crittenden Jr., M.L. Sorensen, Fission-track ages of apatite in the Wasatch Mountains, Utah: an uplift study, Geol. Soc. Am. Mem. 157 (1983) 29–36. [10] J.R. Arnold, W.F. Libby, Radiocarbon dates, Science 113 (1951) 111–120. [11] W.F. Libby, Radiocarbon Dating, University of Chicago Press, Chicago, IL, 1952. [12] R. Berger, A.G. Horney, W.F. Libby, Radiocarbon dating of bone and shell from their organic components, Science 144 (1964) 999–1001. [13] R. Longin, New method of collagen extraction for radiocarbon dating, Nature 230 (1971) 241–242. [14] T.A. Brown, D.E. Nelson, J.S. Vogel, J.R. Southon, Improved collagen extraction by modified Longin method, Radiocarbon 30 (1988) 171–177. [15] T.Y. Ho, L.F. Marcus, R. Berger, Radiocarbon dating of petroleum-impregnated bone from Tar Pits at Rancho La Brea, California, Science 164 (1969) 1051–1952. [16] T.W. Stafford Jr., R.C. Duhamel, C.V. Haynes Jr., K. Brendel, Isolation of proline and hydroxyproline from fossil bone, Life Sci. 31 (1982) 931–938. [17] R.A. Gillespie, R.E.M. Hedges, J.O. Wand, Radiocarbon dating of bone by accelerator mass spectrometry, J. Archaeol. Sci. 11 (1984) 165–170. [18] J.S.O. McCullagh, A. Marom, R.E.M. Hedges, Radiocarbon dating of individual amino acids from archaeological bone collagen, Radiocarbon 52 (2010) 620–634. [19] A. Marom, J.S.O. McCullagh, T.F.G. Higham, A.A. Sinitsyn, R.E.M. Hedges, Single amino acid radiocarbon dating of Upper Paleolithic modem humans, Proc. Natl. Acad. Sci. USA 109 (2012) 6878–6881. [20] D.E. Nelson, A new method for carbon isotope analysis of protein, Science 251 (1991) 552–554. [21] N. Tisnérat-Laborde, H. Valladas, E. Kaltnecker, M. Arnold, AMS radiocarbon dating of bones at LSCE, Radiocarbon 45 (2003) 409–419. [22] M.J. Aitken, Thermoluminescence Dating, Academic Press, London, 1985. [23] M. Ikeya, New Applications of Electron Spin Resonance: Dating, Dosimetry and Microscopy, World Scientific Publishing Co., Singapore, 1993. [24] I. Friedman, R.L. Smith, A new dating method using obsidian: Part I, the development of the method, Am. Antiq. 25 (1960) 476–522. [25] I. Liritzis, N. Laskaris, Fifty years of obsidian hydration dating in archaeology, J. Non Crystalline Solids 357 (2011) 211–219. [26] M.G.L. Baillie, A Slice Through Time. Dendrochronology and Precision Dating, Routledge, Abingdon, UK, 1995.

[27] J. Eighmy, R. Sternberg (Editors), Archaeomagnetic Dating, The University of Arizona Press, Tucson, AR, 1990. [28] R.F. Butler, Paleomagnetism: Magnetic Domains to Geologic Terranes, Blackwell Scientific Publications, Boston, MA, 1992. [29] L.J. Bada, The dating of fossil bones using the racemization of isoleucine, Earth Planet. Sc. Lett. 15 (1972) 223–231. [30] R.E. Taylor, Dating techniques in archaeology and paleoanthropology, Anal. Chem. 59 (1987) 317A–331A. [31] L. Campanella, A. Antonelli, G. Favero, M. Tomassetti, New archeometric method for wood based on an enzymatic biosensor, Actualité chimique 10 (2001) 14–21. [32] L. Campanella, F. Chicco, G. Favero, T. Gatta, M. Tomassetti, Further applications of a new biosensor method for dating cellulosic finds, Ann. Chim. 95 (2005) 133–141. [33] D.A. Kouznetsov, Biochemical methods in cultural heritage conservation studies: an alkylation enzyme, S-adenosylmethionine transmethylase, Stud. Conserv. 45 (2000) 117–126. [34] L. Campanella, T. Carnazza, G. Favero, D. Ferro, M. Tomassetti, Two archaeometric methods for cellulosic textile finds using enzymatic test, in: T.S. Akasheh (Editor), The First International Conference on Science and Technology in Archaeology and Conservation, Fundacion El Legado Andalusi, Granada, Spain, 2002, pp. 117–123. [35] L. Campanella, T. Carnazza, G. Favero, M. Tomassetti, Validazione di un metodo enzimatico di datazione di materiali cellulosici, in: C.D. Amico (Editor), Atti del IV congresso nazionale di archeometria, Patron Editore, Bologna, Italy, 2005, pp. 341–347. [36] L. Campanella, C. Costanza, M. Tomassetti, Determination of environmental persistence and thermogravimetric analysis of paper artificially aged by photoirradiation, Ann. Chim. 96 (2006) 575–585. [37] L. Campanella, C. Costanza, M. Tomassetti, Fotosensore per la datazione di campioni di legno di noce (confronto con i metodi biosensoristico e termogravimetrico), in: C.D. Amico (Editor), Atti del IV congresso nazionale AIAR, Patron Editore, Bologna, Italy, 2007, pp. 653–661. [38] L. Campanella, C. Costanza, M. Tomassetti, Trattamento della carta con antiossidanti. Protezione dai processi di invecchiamento, La Chimica e l’Industria 87 (2005) 84–89. [39] L. Campanella, T. Bevilacqua, C. Costanza, M. Tomassetti, Misura sperimentale della capacità antiossidante della lignina mediante voltammetria ciclica, in: F. Palla (Editor), Atti del convegno AIAr Sistemi biologici e beni culturali, Priulla, Palermo, Italy, 2009, pp. 189–199. [40] S. Reich, G. Leitus, S. Shalev, Measurement of corrosion content of archaeological lead artifacts by their Meissner response in the superconducting state; a new dating method, New J. Phys. 5 (2003) 99, 1-99.9. [41] F. Scholz, U. Schröder, R. Gulaboski, A. Doménech-Carbó, Electrochemistry of Immobilized Particles and Droplets, 2nd ed., Springer, Heidelberg, 2014. [42] A. Doménech-Carbó, J. Labuda, F. Scholz, Electroanalytical chemistry for the analysis of solids: characterization and classification (technical report), Pure Appl. Chem. 85 (2013) 609–631. [43] A. Doménech-Carbó, M.T. Doménech-Carbó, M.A. Peiró-Ronda, Dating archaeological lead artifacts from measurement of the corrosion content using the voltammetry of microparticles, Anal. Chem. 83 (2011) 5639–5644. [44] A. Doménech-Carbó, M.T. Doménech-Carbó, M.A. Peiró-Ronda, I. Martínez-Lázaro, J. Barrio, Application of the voltammetry of microparticles for dating archaeological lead using polarization curves and electrochemical impedance spectroscopy, J. Solid State Electrochem. 16 (2012) 2349–2356. [45] A. Doménech-Carbó, M.T. Doménech-Carbó, S. Capelo, T. Pasíes, I. Martínez-Lázaro, Dating archaeological copper/bronze artifacts using the voltammetry of microparticles, Angew. Chem. Int. Ed. 53 (2014) 9262–9266. [46] A. Doménech-Carbó, M.T. Doménech-Carbó, Y. Lee, L. Osete-Cortina, Potential application of voltammetry of microparticles for dating porcine blood-based binding media used in Taiwanese architectural polychromies, Chem. Asian J. 7 (2012) 2268–2273. [47] F. Scholz, U. Schröder, S. Meyer, K.Z. Brainina, N.F. Zakhchuk, N.V. Sobolev, et al., The electrochenfical response of radiation defects of non-conducting materials. An electrochemical access to age determinations, J. Electroanal. Chem. 385 (1995) 139–142. [48] F. Marini, M. Tomassetti, S. Vecchio, Detailed kinetic and chemometric study of the cellulose thermal breakdown in artificially aged and non aged commercial paper. Different methods for computing activation energy as an assessment model in archaeometric applications, Chem. Centr. J. 6 (Suppl. 2) (2012) 57. [49] L. Campanella, G. Favero, F. Rodante, M. Tomassetti, S. Vecchio, Thermogravimetric and kinetic methods to date wood finds. First results, Ann. Chim. 93 (2003) 897–907. [50] M. Tomassetti, S. Vecchio, C. Costanza, L. Campanella, Possible archaeometric applications deriving from a kinetic study of thermal degradation of cellulose in non aged and artificially aged “Fabriano” paper samples, in: La chimica italiana nell’anno del centenario della Società Chimica Italiana, Italian Chemical Society, Rome, Italy, 2010, pp. 294–299. [51] L. Campanella, C. Costanza, M. Tomassetti, Studio dell’invecchiamento fotochimico della carta, a fini archeometrici, in: C.D. Amico (Editor), Atti del convegno di Caserta dell’Associazione Nazionale di Archeometria, Patron Editore, Bologna, Italy, 2006, pp. 519–527. [52] L. Campanella, C. Costanza, M. Tomassetti, S. Vecchio, Kinetic processing of thermoanalytical curves of cellulose samples and finds. Assessment of possible archaeometric applications, Curr. Anal. Chem. 6 (2010) 260–268. [53] H.G. Wiedemann, G. Bayer, Papyrus: the paper of ancient Egypt, Anal. Chem. 55 (1983) 1220A–1230A.

Please cite this article in press as: Mauro Tomassetti, Federico Marini, Remo Bucci, Luigi Campanella, A survey on innovative dating methods in archaeometry with focus on fossil bones, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.11.017

ARTICLE IN PRESS M. Tomassetti et al./Trends in Analytical Chemistry ■■ (2015) ■■–■■

[54] H.G. Wiedemann, Thermal analysis of the wood components in relation to the growth period of a tree, Thermoch. Acta 229 (1993) 215–228. [55] F. Preußer, Differentialthermoanalyse – eine neue methode zur gemäldeuntersuchung, Maltechnik Restauro 85 (1979) 54–60. [56] M. Odlyha, A. Burmester, Preliminary investigations of the binding media of paintings by differential thermal analysis, J. Thermal Anal. 33 (1988) 1041–1052. [57] M. Odlyha, N.S. Cohen, G.M. Foster, Dosimetry of paintings: determination of the degree of chemical change in museum exposed test paintings (smalt tempera) by thermal analysis, Thermoch. Acta 365 (2000) 35–44. [58] N.S. Cohen, M. Odlyha, R. Campana, G.M. Foster, Dosimetry of paintings: determination of the degree of chemical change in museum exposed test paintings (lead white tempera) by thermal analysis and infrared spectroscopy, Thermoch. Acta 365 (2000) 45–52. [59] M. Odlyha, N.S. Cohen, G.M. Foster, R.H. West, Dosimetry of paintings: determination of the degree of chemical change in museum exposed test paintings (azurite tempera) by thermal and spectroscopic analysis, Thermoch. Acta 365 (2000) 53–63. [60] E. Zuckerkandl, L.B. Pauling, Molecular disease, evolution, and genic heterogeneity, in: M. Kasha, B. Pullman (Editors), Horizons in Biochemistry, Academic Press, New York, NY, 1962, pp. 189–225. [61] M. Kimura, Evolutionary rate at the molecular level, Nature 217 (1968) 624–626. [62] F.J. Ayala, Molecular clock mirages, Bioessays 21 (1999) 71–75. [63] S.Y.W. Ho, M.J. Phillips, A. Cooper, A.J. Drummond, Time dependency of molecular rate estimates and systematic overestimation of recent divergence times, Mol. Biol. Evol. 22 (2005) 1561–1568. [64] S. Horvath, DNA methylation age of human tissues and cell types, Genome Biol. 14 (2013) R115. [65] A.J. Drummond, S.Y.W. Ho, M.J. Phillips, A. Rambaut, Relaxed phylogenetics and dating with confidence, PLoS Biol. 4 (2006) e88. [66] J.H. Schwartz, B. Maresca, Do molecular clocks run at all? A critique of molecular systematics, Biol. Theor. 1 (2006) 357–371. [67] L. Campanella, E. Gregori, M. Tomassetti, G. Visco, Identification of different types of imperial age marble finds using instrumental chemical analysis and pattern recognition analysis, Ann. Chem. 91 (2001) 701–718. [68] M. Tomassetti, L. Campanella, P. Flamini, G. Bandini, Thermal analysis of fictile votive statues of 3rd century B.C, Thermochim. Acta 291 (1997) 117–130 (vedi [8]). [69] L. Campanella, G. Favero, P. Flamini, M. Tomassetti, Prehistoric terracottas from the Lybian Tadrart Acacus, J. Therm. Anal. Calorim. 73 (2003) 127–142. [70] T. Gatta, L. Campanella, C. Coluzza, V. Mambro, P. Postorino, M. Tomassetti, et al., Characterization of black pigment used in 30 BC fresco wall paint using instrumental methods and chemometry, Chem. Centr. J. 6 (Suppl. 2) (2012) 52. [71] M. Tomassetti, F. Marini, L. Campanella, M. Positano, F. Marinucci, Suitable classification of mortars from ancient Roman and Renaissance frescoes using thermal analysis and chemometrics, Chem. Centr. J. 9 (2015) 23. [72] M. Tomassetti, F. Marini, L. Campanella, A. Coppa, Study of modern or ancient collagen and human fossil bones from an archaeological site of middle Nile by thermal analysis and chemometrics, Microch. J. 108 (2013) 7–13. [73] M. Tomassetti, F. Marini, L. Campanella, A. Coppa, Archaeometric classification of ancient human fossil bones, with particular attention to their carbonate content, using chemometrics, thermogravimetry and ICP emission, Chem. Centr. J. 8 (2014) 26. [74] R.F. Heizer, On the methods of Chemical Analysis of Bone as an Aid to Prehistoric Culture Chronology, University of California Archaeological Survey, Berkeley, CA, 1950, pp. 10–14. [75] R.F. Heizer, Long Range Dating in Archaeology, in: A.L. Kroeber (Editor), Anthropology Today, University of Chicago Press, Chicago, IL, 1952, pp. 3–42.

9

[76] A. Carnot, Recherches sur la composition generale et la teneur en fluor des os modernes et des os fossiles de differents ages, Ann. Mines (9, Mem.) 3 (1893) 155–195. [77] H. Barber, Untersuchungen über die Chemische Veranderung von Knochen bei der Fossilization, Palaeobiologica 7 (1939) 217–235. [78] S.F. Cook, The Fossilization of human bone: calcium, phosphate, and carbonate, Univ. Calif. Pub. Am. Arch. Eth. 40 (6) (1951) 263–280. [79] S.F. Cook, R.F. Heizer, The present status of chemical methods for dating prehistoric bone, Am. Antiquity 18 (1953) 354–358. [80] K.P. Oakley, The fluorine-dating method, Yearb. Phys. Anthropol. 5 (1951) 44– 52. [81] R.F. Heizer, S.F. Cook, Fluorine and other chemical tests of some North American human and animal bones, Am. J. Phys. Anthropol. 10 (1952) 289–304. [82] H.C. Ezra, S.F. Cook, H.A. Leon, Current status of the fluorine method of age determination, Kroeber Antrhopological Society Papers 19 (1958) 17–19. [83] F.W. Neuman, B.J. Mulryan, The surface chemistry of bone II: fluoride deposition, J. Biol. Chem. 187 (1950) 655–661. [84] M.G. Lamoureux, The various dating techniques available to archaeologists, 2009. Available from: http://www.sourcinginnovation.com/archaeology/Arch08.htm (accessed October 2015). [85] P. Pin, Contribution de la Biochimie a l’étude des os Préhistoriques, Bulletins et Mémoires de la Société d’Anthropologie de Paris, Serie X 1 (1950) 137–138. [86] T. Thunberg, The citric acid content of older, especially medieval and prehistoric bone material, Acta Physiol. Scand. 14 (1947) 244–247. [87] A. Bartsiokas, A.P. Middleton, Characterization and dating of recent and fossil bone by x-ray diffraction, J. Archaeol. Sci. 19 (1992) 63–72. [88] L. Capasso, R.D. Anastasio, E. Michetti, The use of confocal microscope in the study of ancient bones, Anthropologie 39 (2–3) (2011) 181–186. [89] A. Olcott Marshall, C.P. Marshall, Vibrational spectroscopy of fossils, Palaeontology 58 (2015) 201–211. [90] D. Vincke, R. Miller, É. Stassart, M. Otte, P. Dardenne, M. Collins, et al., Pierna, Analysis of collagen preservation in bones recovered in archaeological contexts using NIR Hyperspectral Imaging, Talanta 125 (2014) 181–188. [91] G. Szoor, Fossil age determination by thermal analysis, J. Therm. Anal. 23 (1982) 83–91. [92] L.F. Lozano, M.A. Pena-Rico, A. Heredia, A. Ocotlan-Flores, A. Gomez-Cortes, R. Velazquez, et al., Thermal analysis study of human bone, J. Mater. Sci. 38 (2003) 4777–4782. [93] T. Sakae, B. Mishima, Y. Kozawa, R.Z. Legeros, Thermal stability of mineralized and demineralized dentin: a differential scanning calorimetric study, Connect. Tissue Res. 33 (1–3) (1995) 193–196. [94] E. Bonucci, G. Graziani, Comparative thermogravimetric, x-ray diffraction and electron microscope investigation of burnt bones from recent, ancient and prehistoric age, Atti Accad. Naz. Lincei Cl. Sci. Fis. Mat. Nat. Rend. 59 (1975) 517–532. [95] H. Haas, J. Banewicz, Radiocarbon dating of bone apatite using thermal release of CO2, Radiocarbon 22 (1980) 537–544. [96] L.D. Mkukuma, J.M.S. Skakle, I.R. Gibson, C.T. Imrie, R.M. Aspden, D.W.L. Hukins, Effect of the proportion of organic material in bone on thermal decomposition of bone mineral: an investigation of a variety of bones from different species using thermogravimetric analysis coupled to mass-spectrometry, high temperature X-ray diffraction, and Fourier-transform infrared spectroscopy, Calcif. Tissue Int. 75 (2004) 321–328. [97] M. Tomassetti, F. Marini, R. Bucci, A. Coppa, L. Campanella, Human fossil bones: archaeometric classification using chemometrics and thermogravimetry. Influence of skeleton fossilization and its anatomical parts, Microch. J. 124 (2016) 396–401.

Please cite this article in press as: Mauro Tomassetti, Federico Marini, Remo Bucci, Luigi Campanella, A survey on innovative dating methods in archaeometry with focus on fossil bones, Trends in Analytical Chemistry (2015), doi: 10.1016/j.trac.2015.11.017