Journal of Archaeological Science 37 (2010) 3022e3029

Contents lists available at ScienceDirect

Journal of Archaeological Science

journal homepage: http://www.elsevier.com/locate/jas

Distinguishing between calcites formed by different mechanisms using infrared

spectrometry: archaeological applications
Lior Regev a, b, *, Kristin M. Poduska c, Lia Addadi a, Steve Weiner a, b, Elisabetta Boaretto d, e
a
Department of Structural Biology, Weizmann Institute of Science, Rehovot 76100, Israel
b
Kimmel Center for Archaeological Science, Weizmann Institute of Science, Rehovot 76100, Israel
c
Department of Physics and Physical Oceanography, Memorial University of Newfoundland, St. John’s, NL A1B 3X7, Canada
d
Radiocarbon and Cosmogenic Isotopes Laboratory, Kimmel Center for Archaeological Science, Weizmann Institute of Science, Rehovot 76100, Israel
e
Department of Land of Israel Studies and Archaeology, Bar-Ilan University, Ramat-Gan 52900, Israel

a r t i c l e i n f o a b s t r a c t

Article history: Infrared spectrometry is a well-established method for the identification of minerals. Due to its
Received 11 February 2010 simplicity and the short time required to obtain a result, it can be practiced on-site during excavation
Received in revised form using portable infrared spectrometers. However, the identification of a mineral may not be sufficient. For
8 June 2010
example, a lime plaster floor and a crushed chalk surface have a similar appearance and are composed of
Accepted 23 June 2010
the same mineral e calcite. Here we exploit differences in the infrared spectra of geogenic, biogenic and
pyrogenic calcites for the identification of each calcite type. The infrared calcite spectrum has three
Keywords:
characteristic peaks in the region of 400e4000 cm
1, designated n2, n3, and n4. When a calcite sample is
Calcium carbonate
Infrared spectroscopy
ground, as part of the measurement preparation procedure, some grinding dependent changes will be
Pyrotechnology revealed in the infrared spectrum. With additional grinding, the n3 peak narrows and the heights of the
Diagenesis n2 and n4 peaks decrease, when both are normalized to the n3 height. By plotting the normalized heights
Ash of the n2 versus the n4 of several grindings of the same sample, a characteristic trend line is formed for
Plaster each calcite type. The trend lines of geogenic calcites have the shallowest slopes and highest n4 values
when compared to pyrogenic calcites, which can be further divided to ash and plaster/mortar samples.
This method can assist in distinguishing between the various calcites, and provide insights into homo-
geneity and preservation state of the calcitic materials in question.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction intimately mixed with the local sediment that may contain geo-
genic calcite. The identification of the different forms of calcite in an
Calcite is a common mineral in geological and archaeological archaeological site can provide information on the activities per-
contexts. Calcite is the main component of limestone, chalk, formed in different areas of the site, such as various pyrotechno-
marble, travertine, and is often a major component of wind-borne logical activities, and can shed light on aspects of site formation
sediment (loess) and cemented sediments. Calcite is also a common processes. Here we demonstrate how Fourier transform infrared
component of invertebrate shells and other skeletal elements, and spectrometry (FTIR) can be used to distinguish between calcites
is produced anthropogenically in the form of wood ash, as well as formed by different mechanisms. We illustrate the use of the
plaster and mortar. All of these forms of calcite can be found in approach by analyzing the various calcite components of an ancient
archaeological sites, sometimes in considerable quantities. plaster.
Although the identification of the various calcite types present in Geogenic, biogenic and pyrogenic calcites are formed by
archaeological sites is sometimes straightforward, difficulties may different mechanisms, and this in turn can influence characteristics
arise. For instance, a white crushed chalk surface can be mistakenly such as crystallinity, coherence length, domain size, and crystal
identified as a lime plaster floor. Fine-grained ash calcite is often habit. Large geogenic calcite crystals, with well-defined crystal
difficult to identify, as it can undergo diagenetic changes, and/or be habits, are presumed to form slowly from saturated solutions. In
contrast, many biogenic calcites are thought to form via an amor-
phous calcium carbonate precursor phase that crystallizes in vivo
* Corresponding author at: Kimmel Center for Archaeological Science, Weizmann
(Addadi et al., 2003; Beniash et al., 1997; Gago-Duport et al., 2008;
Institute of Science, Rehovot 76100, Israel. Tel.: þ972 8 934 6275; fax: þ972 8 934
6062. Politi et al., 2008). Chalk and limestone typically originate as
E-mail address: lior.regev@weizmann.ac.il (L. Regev). biogenic calcite, but they have often undergone dissolution and

0305-4403/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jas.2010.06.027
 L. Regev et al. / Journal of Archaeological Science 37 (2010) 3022e3029 3023

reprecipitation processes to different extents (Pettijohn, 1957). The absorption peak narrows with increased grinding of the sample as
two main forms of pyrogenic calcites found in archaeological sites the n2/n4 ratio increases.
are wood ash and lime plaster/mortar, each with a different Here we present a new approach for using the infrared spectra
formation pathway. Lime plaster or mortar is formed when lime- of calcite for distinguishing between calcites formed by different
stone or chalk is heated near 750  C, the point at which the calcium processes. We show that the relative IR peak heights change with
carbonate decomposes to calcium oxide (CaO) and carbon dioxide grinding in a consistent way in a single material, and that calcites of
(Boynton, 1980, pp. 159, 162). Water is then added to form hydrated different origins have different relative peak heights. This allows us
lime (Ca(OH)2). The slaked lime is often mixed with other additives to more reliably distinguish between geogenic, biogenic, and
to increase volume and strength. These additives are frequently in pyrogenic calcites, and in addition does not require that all samples
the form of geogenic calcite (Borrelli, 1999). When the paste is set as be ground to the same extent. This approach can be used for rapid
surfaces or a binding material, it dehydrates and absorbs carbon sample analysis in the field.
dioxide from the atmosphere to reform as calcite. Wood ash is
usually formed from calcium oxalate crystals (CaC2O4) found in the 2. Materials and methods
wood. When heated to around 500  C, the calcium oxalate
decomposes to calcite (CaCO3) by the release of carbon monoxide 2.1. Materials
(CO) (Dollimore, 1987; Huaqing, 1989). Wood ash calcite may also
be exposed to temperatures above 750  C in kilns and ovens used Geogenic samples were collected in Israel. Limestones were
for various pyrotechnological activities. This ash will decompose to collected from coastal Tel Dor (20 km south of Haifa, see Sharon
CaO and reform as calcite in the same way as lime plaster forms et al., 2009, designated as ‘Limestone A’ in the paper) and from
(Canti, 2003; Dollimore, 1987; Huaqing, 1989). Tel Tsafit (Tel es-Safi/Gath, 20 km south-east of Ashdod, see Maeir,
Infrared spectroscopy is a well-established method for identi- in press, ‘Limestone B’). Chalk samples were collected from Ramat
fying different mineral and organic components in archaeological Rachel archaeological site close to Jerusalem (Lipschits et al., 2009,
samples. These material-specific IR “fingerprints” are possible ‘Chalk A’), and from Tel Tsafit (‘Chalk B’). The chalky limestone
because the peaks in an infrared spectrum are the result of the sample was collected from the quarry of Nesher Israel Cement
absorption of photons at specific energies corresponding to vibra- Enterprises Ltd. in Ramla (designated as ‘chalky limestone’).
tions or rotations of molecules or of specific chemical bonds within Geological calcite crystals (Iceland Spar) were from Mexico and
them. Any change in the character of these bonds will change the were purchased from Ward’s (catalog no. 49H1600). Sea urchins
characteristic energy of their vibrations and can result in a change were supplied by the Israel Oceanographic and Limnological
in the peak shape, width, height, and/or position. Particularly Research Company, Eilat, Israel. All fresh plasters were made by
relevant to this study is that the vibrational and rotational infrared calcination and slaking of the chalky limestone sample. Heating
spectra are sensitive to the molecular environment, both crystalline temperatures and times are listed in Table 1. Modern plasters A, C
and amorphous. In addition, it is well documented that differences and D were slaked by placing ca. 30 g of the calcined sample in
in particle sizes and morphologies can also affect the sizes and a 400 ml beaker, and adding an excess amount of water one day
shapes of IR absorption peaks (Duyckaerts, 1959; Lane, 1999; after the heating. A layer of 1e3 cm of plaster was formed in each
Ruppin and Englman, 1970). An infrared spectrum is therefore beaker. Modern plaster B was slaked by the addition of excess water
affected by order within a material, both at the local atomic level while the calcined rock was still hot in a cooking pot. A plaster layer
and at longer length scales. 1e3 mm thick was formed. All plasters were left on the laboratory
Calcite has three characteristic IR absorption peaks in the range shelf for recarbonation at room temperature of 20e25  C, without
between 400 and 4000 wavenumbers (measured in cm
1 units, monitoring the environmental conditions. The recarbonating
which is the reciprocal of wavelength): n3 at 1420 cm
1, n2 at samples were monitored by infrared analysis, detecting the calcite
874 cm
1, and n4 at 713 cm
1. All of these absorptions are influ- formation and the decrease in the calcium hydroxide peak intensity
enced by the different ways in which the CeO bonds of the (narrow, clear peak at 3643 cm
1). The Yiftahel plaster was
carbonate unit interact with the infrared radiation. It has been collected from one of the floors of the pre-pottery Neolithic B strata
shown experimentally (Addadi et al., 2003; Aizenberg et al., 1997; on the site of Yiftahel (8 km north-west of Nazareth, lower Galilee,
Beniash et al., 1997; Politi et al., 2004) and theoretically (Gueta see Khalaily et al., 2008. The sample was taken from Building 501,
et al., 2007) in biogenic calcite from the sea urchin Paracentrotus square G17, L5098, and registered as B50183). The preparation
lividus, that the ratio of the n2/n4 peak heights decreases as the modes of the modern ashes are listed in Table 1. Ash 900  C was left
initially deposited amorphous calcium carbonate crystallizes into after heating in a box on the shelf for air slaking and recarbonation.
calcite. Chu et al. (2008) used the n2/n4 peak heights ratio to Sample depth did not exceed 3 mm. The archaeological ash was
distinguish between geogenic and pyrogenic calcites in archaeo- collected from an Iron Age layer at the site of Tel es-Safi/Gath (Area
logical contexts. The same study also showed that the calcite n3 A, L111222).

Table 1
Preparation modes of modern plaster and ash samples.

Sample Lab number Raw material Heating device Temperature Duration Date

Modern plaster A DS2III1 Chalky limestone Open bonfire 965 C (average) 3 h 1.2.07

Modern plaster B DS1III2pot Chalky limestone Open bonfire 980 C (average) 3 h 1.2.07

Modern plaster C Nesher 29.1.08 Chalky limestone Lab oven 900 C 1 h 29.1.08

Modern plaster D DS1III3 Chalky limestone Open bonfire 980 C (average) 3 h 1.2.07

Modern ash A R2 bag 1 Various wood types Open bonfire 805 C (average) 8 h 31.1.07

Modern ash B SGA10 Various wood types Open bonfire 620 C (average) 3 h 21.7.07

Modern ash C Weizmann Pine, Pine bark Lab oven 500 C 2 h 26.7.09
rings 5, 17e19
Modern ash from RT5911 Maerua Crassifolia from the Lab oven 500  C 1h 8.12.09
ancient wood Cave of Letters site, Israel
Ash 900  C R2 bag 1, 900 Modern ash A Lab oven 900 C
2h 31.8.08
3024 L. Regev et al. / Journal of Archaeological Science 37 (2010) 3022e3029

Fig. 1. (a) Infrared spectra of calcite, ground to different extents. The sample measured is an Iceland Spar crystal. I e Lightly ground sample. The designation of the peaks as n3, n2
and n4 is marked, and baselines are shown for the measurements of the n2 and n4 heights. II e Heavily ground sample. The baseline used for the height measurement of the n3 peak
is shown. The length (in wavenumbers) of the horizontal line at the half height of the n3 peak is the full width at half maximal height value (FWHM). The FWHM values of the
presented peaks are: Spectrum I e n3 ¼ 127 cm
1, n2 ¼ 14 cm
1, n4 ¼ 7.4 cm
1. Spectrum II e n3 ¼ 89 cm
1, n2 ¼ 10.6 cm
1, n4 ¼ 7.0 cm
1. The n2/n4 heights ratios are: Spectrum I e
2.1. Spectrum II e 3.8. (b) Plot of the n2 versus n4 peak heights of geogenic and pyrogenic calcites after each spectrum was normalized to the corresponding n3 peak height. For each
type of calcite, data points correspond to successive grindings of the same specimen. The values along the trend line for the sparry calcite crystal sample designate the width
(FWHM) of the n3 peak (n.a.u. e normalized absorbance units).

2.2. Fourier transform infrared spectrometry (FTIR) 2.3. Radiocarbon dating of geogenic calcites

A few milligrams of sample were homogenized and powdered in 20e30 g of each sample were crushed and dissolved in 5N HCl
an agate mortar and pestle. About 0.3 mg were left in the mortar for CO2 extraction. The gas was chemically transformed (through
and mixed with about 40 mg of KBr and pressed into a 7 mm pellet C2H2 and Li2C2) into benzene for dating using a liquid scintillation
using a manual hydraulic press (Specac). Each sample was counter at the Radiocarbon and Cosmogenic Isotopes Laboratory,
measured repeatedly, either by repetitive grinding of the same KBr Kimmel Center of Archaeological Science, Weizmann Institute of
pellet or by using different aliquots of the same sample that were Science, Rehovot, Israel.
ground to different extents. Both methods produced the same
results. All repetitive grindings were made by hand with an agate 3. Results
mortar and pestle. Typically, a few seconds of regrinding were
applied. Infrared spectra were obtained at 4 cm
1 resolution for 32 FTIR analysis requires powdered samples, and it has been shown
scans using a Nicolet 380 instrument (Thermo). The baselines for that differences in the amount of grinding affect peak widths and
the heights measurements of the n3, n2 and n4 peaks were deter- relative heights in different archaeological materials, including
mined following the method of Chu et al. (2008). The n2 and n4 apatites and calcites (Chu et al., 2008; Surovell and Stiner, 2001).
heights were normalized to a n3 height of 1000, corresponding to However, we find that by normalizing the n2 and n4 peak heights
1.0 absorbance unit. relative to the height of the n3 peak, a consistent trend emerges that
 L. Regev et al. / Journal of Archaeological Science 37 (2010) 3022e3029 3025

Sea Urchin Test

Sea Urchin Secondary Spine
Sea Urchin Primary Spine
Iceland Spar trend line R2 = 0.9405
Sea Urchin Secondary Spine trend line
470 Sea Urchin Primary Spine trend line
R2 = 0.9535
Sea Urchin Test trend line
[n.a.u]
420
R2 = 0.9862
ν2 Height

370

320

270
30 50 70 90 110 130 150 170 190 210 230
ν4 Height [n.a.u]
Fig. 2. Plot of the n2 versus n4 peak heights of various biogenic calcites. The black line corresponds to the sparry calcite crystal polynomial trend line from Fig. 1 (n.a.u. e normalized
absorbance units).

decouples the effect of grinding from the intrinsic order in the 3.1. Biogenic calcite
material. Calcite infrared spectra of an Iceland Spar crystal ground
to different extents is shown in Fig. 1a. It can be seen that the lightly In order to gain more insight into the significance of the changes
ground sample (spectrum I) has wider peaks and a smaller n2/n4 in Fig. 1, we analyzed primary and secondary sea urchin spines, as
height ratio than the heavily ground one (spectrum II). The widths well as a sea urchin test (body skeleton). Although the slopes of the
of the peaks are measured by their full widths at half maximum three sea urchin skeletal element trend lines are fairly similar,
(FWHM) values (see Fig. 1a for illustration). The four calcite samples significant separation between the trend lines is evident (Fig. 2).
in Fig. 1b have distinctly different trend lines for the normalized n2 The trend line for the secondary spine is closer to that of the single
versus n4 peak heights. With increased grinding, the relative n2 and sparry calcite crystal as compared to the primary spines. An earlier
n4 peak heights decrease, and all peaks in the spectra narrow. As an study indeed found textural differences between spines at different
example, the n3 peak widths (FWHM) are included for the Iceland maturation levels, using X-ray diffraction (Aizenberg et al., 1997).
Spar calcite in Fig. 1b. Since trend lines are produced, no special We note that the location of the peak maximum of the n4 peak
attention to the exact degree of repetitive grinding is required. shifts from 713 cm
1 to around 717 cm
1 for the sea urchin test,
The different trend lines in Fig. 1b show that there is a system- which reflects increasing magnesium incorporation into the calcite
atic difference in the n2 versus n4 peak heights for calcites that were structure (Bischoff et al., 1985; Falini et al., 1998; Farmer, 1974).
formed in different ways. The shallowest slope corresponds to
a geogenic single crystal of Iceland Spar, which is known to have 3.2. Calcitic ash
high atomic order over macroscopic length scales (Bragg, 1914;
Gorton, 1916), while the steepest slope corresponds to a modern The formation of ash calcite is unique. At around 500  C, calcium
lime plaster. Modern ash and geogenic limestone fall in between, oxalate (CaC2O4) in the wood and bark degrades, loses carbon
with the limestone being more similar to the geogenic Iceland Spar monoxide (CO) and becomes calcite (CaCO3) (Dollimore, 1987;
sample. Huaqing, 1989). If the ash temperature exceeds ca. 750  C, the

Modern Ash A
Modern Ash B
Modern Ash C
Modern Ash from Ancient Wood
485 Archaeological Ash
Ash 900C
Iceland Spar trend line
435
ν2 Height [n.a.u]

385

335

285

235
10 60 110 160 210
ν4 Height [n.a.u]
Fig. 3. n2 and n4 peak heights of various ash calcites. The black line corresponds to the sparry calcite polynomial trend line from Fig. 1. Modern ashes A, B, C were produced from
modern woods (see Table 1 for details on all modern ashes). Modern ash from ancient wood was made by burning a 2000-year-old wood from the Cave of Letters (Dead sea area,
Israel). The archaeological ash is a 3000-year-old ash sample from Tel es-Safi/Gath (n.a.u. e normalized absorbance units).
3026 L. Regev et al. / Journal of Archaeological Science 37 (2010) 3022e3029

650 Modern Plaster A

Modern Plaster B
600 Modern Plaster C
Modern Plaster D
Yiftahel top
550 Yiftahel white nodules
Yiftahel yellow
ν2 Height [n.a.u] 500 Limestone A
Iceland Spar trend line
450

400

350

300

250
35 55 75 95 115 135 155 175
ν4 Height [n.a.u]

Fig. 4. n2 and n4 peak heights of various plaster calcites. The black line shows the sparry calcite single crystal polynomial trend line from Fig. 1. The modern plaster preparation
modes appear in Table 1. The archaeological samples fall in the grey area, between the geogenic raw materials and the fresh plasters (n.a.u. e normalized absorbance units).

calcite will decompose to CaO and CO2, followed by water limestones (Fig. 6, ‘Limestone B’, from Tel Tsafit), a relatively well-
absorption and recarbonation back to form calcite (Boynton, 1980). consolidated chalky limestone (from Ramle, Israel) and a chalk
Fig. 3 is a plot of a variety of different modern ash calcites produced derived from a depth of 3 m below the surface at Ramat Rahel
in the oven at 500  C or in bonfires. Although a small amount of (Fig. 6, ‘Chalk A,
3 m’), have trend lines close to that of the sparry
the bonfires ‘low-temperature’ ash decomposed to CaO, as in ‘high- single calcite crystal. Limestone A and especially the other two
temperature’ ash (based on the presence of the Ca(OH)2 infrared chalk samples derived from close to the extant ground surface
absorption peak at 3643 cm
1), the majority of the calcite was are clearly less like the geogenic Iceland Spar than the first
formed without this additional step. One ash sample (Modern ash three mentioned samples. Radiocarbon and stable carbon isotope
A) was reheated at 900  C. All the samples cluster around the same measurements of the samples (Table 2) show that there was carbon
trend line, except for the one that was reheated to 900  C, which exchange between the rock and the atmosphere in the not so
has a much higher slope and is the furthest away from the sparry distant past, which could lead to a diagenetic imprint on the
calcite trend line. We also note that one of the ash samples is from formation of those chalk and limestone calcites.
an archaeological site and is around 3000 years old. It falls within
the cluster of the modern ash samples. 4. Discussion

3.3. Calcitic plaster 4.1. Comparison of the trend line approach

Fig. 4 shows a plot of the n2 versus n4 heights of four different Here we present a new approach for distinguishing between
modern plaster samples produced in the laboratory, as well as calcite crystals formed by different mechanisms, based on changes
a detailed analysis of calcite samples from a piece of plaster from in the relative heights of their IR absorption peaks. This trend line
the PPNB site of Yiftahel (northern Israel) that is about 9000 years method offers significant advantages over using a single n2/n4
old (Fig. 5). The modern plaster samples all fall on a well-defined height ratio, as we had used in an earlier study, to distinguish
trend line, even though plasters C and D are not fully carbonated
and still contain some calcium hydroxide (based on the infrared
absorption peak at 3643 cm
1). The archaeological plaster samples
are all below this line, but above the trend line for limestone.
Limestone is the starting material for producing plaster. In the
Yiftahel sample, limestone was also mixed into the sample to
increase the bulk and improve its mechanical properties. The Yif-
tahel plaster samples extracted from the thin surface layer all fall on
the same trend line. A sample from the rectangular inclusion (Fig. 5)
also falls on this trend line. This may indicate that the rectangular
piece was from an older plaster floor surface that was reconstructed
and used as aggregates in the sampled floor. Analysis of the yellow
basal layer itself (Fig. 4 crosses) yields a different trend line, closer
to the limestone trend line. The locations of the trend lines
observed in Fig. 5 probably reflect the extent of mixing between the
pyrogenic calcite, prepared by heating the raw material (such as
limestone), and geogenic calcitic aggregates that were added to it.

3.4. Geogenic calcite

Fig. 5. A section of a PPNB plaster floor from the site of Yiftahel. Three layers can be
seen: a top thin surface layer (A), a second red shaded layer (B), and a lower basal
Fig. 6 shows n2 versus n4 peak height plots of limestones and yellow colored layer (C). An asterisk marks a white rectangular piece, similar in color
chalks obtained from various locations in Israel. One of the and texture to the surface layer.
 L. Regev et al. / Journal of Archaeological Science 37 (2010) 3022e3029 3027

650 Limestone A
Limestone B
Chalk A, -1m
600 Chalk A, -3m
Chalk B
550 Chalky Limestone
Modern Plaster A

ν2 Height [n.a.u]
Iceland Spar trend line
500

450

400

350

300

250
40 90 140 190 240
ν4 Height [n.a.u]
Fig. 6. n2 and n4 peak heights of various geogenic calcites. The black line corresponds to the sparry calcite single crystal polynomial trend line from Fig. 1. The
1 m and
3 m in the
Chalk A refer to samples derived from 1 m to 3 m below ground surface level (n.a.u. e normalized absorbance units).

between geogenic calcite, ash calcite and plaster calcite (Chu et al., sparry calcite. Biological calcites can also contain magnesium, and
2008). In order to normalize the effects grinding has on peak this has an effect on the n4 peak (Bischoff et al., 1985; Falini et al.,
heights and widths, Chu et al. (2008) recommended that each 1998; Farmer, 1974). The test contains the most magnesium from
sample be ground until the width of the n3 peak falls in a “window” the three analyzed samples, while the secondary spines have the
between 110 and 130 cm
1. In contrast, the new trend line method lowest amount (Zolotoyabko et al., 2010), and this may also affect
utilizes samples over a wide range of grinding, making the trend the trend lines.
line independent of the degree to which any one sample is ground. The correlation between trend line slope and position with
By analyzing the same sample many times, the changes in the differences in the intrinsic order suggests that our IR-based analysis
infrared spectrum due to grinding can be differentiated from the method could provide insights for future studies into the processes
changes due to differences in the intrinsic order among calcites involved in pyrogenic calcite formation.
formed by different mechanisms. In this way, calcites formed by
different mechanisms, as well as changes in a material due to 4.2. Possible archaeological applications
diagenesis, or mixtures of calcites from different origins, can all be
recognized more reliably. As calcite is a common component of sediments in many
We show that calcites formed by different mechanisms have archaeological sites, this mode of analysis can be used to identify
characteristic trend lines when the normalized heights of the n2 the source of the calcite, namely geogenic, biogenic, and pyrogenic
and n4 peaks are plotted against each other. We observe a correla- in the form of ash or plaster. This differentiation is only possible if
tion between the extent of crystalline order and the slope and the calcite is relatively well preserved, as recrystallization of the
position of these different n2 versus n4 trend lines. A single crystal of calcite will most likely change the atomic order. In fact scattering in
sparry calcite has the lowest slope and is located to the extreme the data points and/or an absence of a well-defined trend line may
right of the v2 versus v4 peak height plot (Fig. 7). These calcite well indicate that the calcite has been altered diagenetically. One
crystals are known to have extremely high atomic order over way to independently assess the preservation state of the calcite is
macroscopic length scales (Bragg, 1914; Gorton, 1916). In contrast, to determine whether gastropod shells or the shells of fresh water
calcites from other origins do not exhibit this same extent of order. bivalves that are in the proximity of the calcite sample under
Although biogenic samples can also be highly crystalline at a local investigation are still composed of 100% aragonite. As aragonite is
atomic level, a detailed synchrotron X-ray diffraction study showed less stable than calcite at ambient temperatures, this is a good
that primary sea urchin spines have perfectly ordered domains that indication that the calcite should also be well preserved (Weiner,
are larger than those of the secondary spines, but smaller than 2010).
those of sparry calcite (Aizenberg et al., 1997). On the other hand, Ash calcite is often produced in large amounts in archaeological
the same study showed that the extent of anisotropy of the sites. The trend lines for low- and high-temperature ash are quite
secondary spines is higher than the primary spines, and similar to different (Fig. 3). The plot can thus be used to identify ash produced
the anisotropy of pure synthetic calcite. We note that trend lines of in high-temperature industries (above ca. 800  C) such as metal
the primary and secondary spines are separated in Fig. 2, and that production or in certain ceramic kilns. Note that even though the
the trend line of the secondary spines is closer to the trend line of high temperature ash and the fresh plasters were calcined and
slaked (although by a different slaking method), each has
Table 2 a different trend line (Fig. 7). This implies that characteristics of the
14
C and d13C values of various geogenic calcites. initial calcite affect the trend line position and slope despite the
Sample pMC d13C calcination and recarbonation processes.
As ash calcite is fine-grained and can easily be dispersed by
Chalk A (
1 m) 12.06  0.11
7.4&
Chalk B 2.08  0.23
1.4& wind and/or trampling, it is often not easy to identify ash especially
Limestone A 2.38  0.07
4.1& when mixed with other sedimentary minerals. The trend line of the
Chalk A (
3 m) 1.17  0.05
6.3& common form of ash produced at 500  C is quite distinct from
Chalky limestone 0.90  0.10 þ1.8& limestone and plaster, but does overlap with chalk. The presence of
Background (marble) 0.65  0.06 þ2.4&
chalk at a site not only complicates the identification of ash, but
3028 L. Regev et al. / Journal of Archaeological Science 37 (2010) 3022e3029

440

ν2 Height [n.a.u]
390

340

Iceland Spar trend line

290 Limestone trend line
Modern Ash A trend line
Modern Plaster A trend line
Ash 900 trend line
240
0 50 100 150 200 250 300
ν4 Height [n.a.u]
Fig. 7. Plot of the n2 versus n4 peak height trend lines of geogenic and pyrogenic calcites. Note that only the linear portions are shown. The trend lines are based on data presented in
Figs. 1 and 3 (n.a.u. e normalized absorbance units).

because chalk itself undergoes diagenesis when close to the sedi- This method is based on differences in dissolution rates between
ment surface, it may not have a consistent order. One interesting the pyrogenic calcite of the binder and the geogenic calcite of the
potential geochemical application is to use this method to identify aggregates. Based on Fig. 6, the diagenetic chalk might have
well-preserved chalk that still has the biogenic order of the original a similar dissolution profile to the pyrogenic calcite, and thus affect
skeletal elements that make up the chalk. This “pristine” chalk may the interpretation of the radiocarbon results. Therefore, analyzing
provide stable isotope and trace element values that can be reliably the infrared spectrum of the plaster/mortar aggregates by the
used for reconstructing past climates at the time the organisms that method proposed here might predict possible difficulties that may
produced the skeletons lived in the ocean. arise during the sequential dissolution procedure.
We demonstrate the usefulness of the method for two archae- It is often very useful to be able to identify the type of calcite
ological samples. Fig. 3 shows that the trend line of a 3000-year-old encountered during an excavation using an infrared spectrometer
archaeological ash sample overlies the trend lines of modern ash on-site. As it is cumbersome to run multiple infrared spectra for
samples. This implies that this archaeological ash is well preserved. each sample, we recommend preparing a “master plot” of the main
Fig. 4 shows the analysis of a 9000-year-old plaster sample, primary materials at the site that includes the common geogenic
compared with modern fresh plasters and limestone; the primary calcites, as well as the trend lines for modern ash and plaster. Such
materials used for preparing this plaster. The thin layer on the a plot is shown in Fig. 7. The individual analyses can then be located
surface of the plaster (Fig. 5) has a trend line close to the fresh on the master plot and preliminary conclusions can be drawn about
plaster. Furthermore all the samples analyzed from the surface their origins of formation. For more detailed analysis, we would
(within a 30  10 cm sample) all fall on the same trend line. This recommend analyzing the same sample after homogenization
shows that the surface plaster is homogenous, and was probably several times with different extents of grinding, rather than
produced from a thin layer of smooth hydrated lime (Ca(OH)2). This regrinding the same KBr pellet, for a better representation of the
trend line however does not overlap with the modern plaster line, original sample material.
indicating that some diagenetic changes have taken place on the
plaster surface resulting in a shift of the trend line towards the 5. Conclusions
sparry calcite trend line. The lower yellow layer in the plaster
contains geogenic calcite aggregates (Fig. 5). These aggregates The normalized n2 versus the n4 peaks height trend lines for
produce a trend line located lower in the plot, closer to the lime- samples ground to different extents is an efficient, rapid, and
stone trend line. Note that both diagenesis and mixing with geo- effective tool to distinguish between well-preserved calcites
genic limestone may result in a lower trend line for a plaster formed by different processes. The basis for this differentiation
sample. correlates with characteristic differences of local and extended
In our analysis of geogenic chalk deposits at Ramat Rahel (Fig. 6, atomic order in the calcite. This analysis method can be applied
‘Chalk A’), we observed that the data points of the chalk close to the with an on-site infrared spectrometer during an excavation, and
present day surface are closer to those of fresh plaster compared enables the identification of different calcite types, as well as
with chalk from the same rock formation located 3 m below the a means of assessing their preservation states.
surface (Fig. 6). This is unexpected, as diagenesis of calcite usually
involves dissolution followed by recrystallization, which should Acknowledgements
result in a calcite product that is similar to geogenic sparry calcite.
This may indicate that in this chalk the diagenetic process involves We would like to thank Eugenia Mintz for help with the radio-
some other mechanism of reorganization of the calcite crystals. We carbon dating, and to all the people who provided us with samples
also observed that these geological chalks produced finite radio- or allowed us to collect samples from the sites under their super-
carbon ages, implying that diagenesis is “ongoing”. When such vision: Dr. Yael Politi for the sea urchin samples; Dr. Dan Shavit
diagenetic chalk is incorporated into plaster or mortar as an from Nesher Cement Enterprises Ltd.; Ido Rasis, Keren Kayemeth
aggregate, it might cause difficulties when dating the plaster/ LeIsrael for supplying the wood for the open bonfires; Shira Gur-
mortar by the sequential dissolution method (Lindroos et al., 2007). Arie for supplying us with Modern ash B and the archaeological ash
 L. Regev et al. / Journal of Archaeological Science 37 (2010) 3022e3029 3029

samples; Prof Richard Freund and Prof. Rami Arav for the ancient Farmer, V.C., 1974. The Infrared Spectra of Minerals. Mineralogical Society, London.
Gago-Duport, L., Briones, M.J.I., Rodríguez, J.B., Covelo, B., 2008. Amorphous
wood from the Cave of Letters; Drs. Hamoudi Khalaily, Ianir Mile-
calcium carbonate biomineralization in the earthworm’s calciferous gland:
vski and Nimrod Getzov from the Yiftahel excavations; Prof. Aren pathways to the formation of crystalline phases. Journal of Structural Biology
Maeir from the Tell es-Safi/Gath Archaeological Project; Prof. Oded 162, 422e435.
Lipschits and Dr. Yuval Gadot from Ramat Rachel excavations; Gorton, W.S., 1916. The X-ray spectrum of tungsten. Physical Review VII (2),
203e208.
Drs. Ilan Sharon and Ayelet Gilboa from Tel Dor excavation; and Gueta, R., Natan, A., Addadi, L., Weiner, S., Refson, K., Kronik, L., 2007. Local atomic
Mr. Ulf Fornhammar for introducing us to calcite calcination order and infrared spectra of biogenic calcite. Angewandte Chemie Interna-
using open bonfires. This research was funded in part by the tional Edition 46, 291e294.
Huaqing, L., 1989. Effect of chemical adsorption of calcium oxalate monohydrate on
European Research Council under the European Community’s DTA baseline. Thermochimica Acta 141, 151e157.
Seventh Framework Programme (FP7/2007-2013)/ERC grant Khalaily, H., Milevski, I., Getzov, N., Hershkovitz, I., Barzilai, O., Yarosevich, A.,
agreement no 229418, and in part by the Kimmel Center for Shlomi, V., Najjar, A., Zidan, O., Smithline, H., Liran, R., 2008. Recent excavations
at the Neolithic site of Yiftahel (Khalet Khalladyiah), Lower Galilee. Neo-Lithics
Archaeological Science, Weizmann Institute of Science. 2/08, 3e11.
Lane, M.D., 1999. Midinfrared optical constants of calcite and their relationship to
particle size effects in thermal emission spectra of granular calcite. Journal of
References Geophysical Research, Planets 104 (E6), 14099e14108.
Lindroos, A., Heinemeier, J., Ringbom, A., Brasken, M., Sveinbjornsdottir, A., 2007.
Addadi, L., Raz, S., Weiner, S., 2003. Taking advantage of disorder: amorphous Mortar dating using AMS C-14 and sequential dissolution: examples from
calcium carbonate and its roles in biomineralization. Advanced Materials 15, medieval, non-hydraulic lime mortars from the Aland Islands, SW Finland.
959e970. Radiocarbon 49 (1), 47e67.
Aizenberg, J., Hanson, J., Koetzle, T.F., Weiner, S., Addadi, L., 1997. Control of Lipschits, O., Oeming, M., Gadot, Y., Arubas, B., 2009. The 2006 and 2007 excavation
macromolecule distribution within synthetic and biogenic single calcite crys- seasons in Ramat Rahel. Israel Exploration Journal 59, 1e20.
tals. Journal of the American Chemical Society 119 (5), 881e886. Maeir, A. (ed.), Tell es-Safi/Gath I: Report on the 1996e2005 Seasons. Ägypten und
Beniash, E., Aizenberg, J., Addadi, L., Weiner, S., 1997. Amorphous calcium carbonate Altes Testament. vol. 69. Wiesbaden: Harrassowitz (in press).
transforms into calcite during sea urchin larval spicule growth. Proceedings of Pettijohn, F., 1957. Sedimentary Rocks, second ed. Harper & Row, New York.
the Royal Society of London Series B-Biological Sciences 264 (1380), 461e465. Politi, Y., Arad, T., Klein, E., Weiner, S., Addadi, L., 2004. Sea urchin spine calcite
Bischoff, W.D., Sharma, S.K., Mackenzie, F.T., 1985. Carbonate ion disorder in forms via a transient amorphous calcium carbonate phase. Science 306 (5699),
synthetic and biogenic magnesian calcites e A Raman spectral study. American 1161e1164.
Mineralogist 70 (5e6), 581e589. Politi, Y., Metzler, R.A., Abrecht, M., Gilbert, B., Wilt, F.H., Sagi, I., Addadi, L.,
Borrelli, E., 1999. Binders e ARC Laboratory Handbook, ICCROM ARC Laboratory Weiner, S., Gilbert, P.U.P.A., 2008. Transformation mechanism of amorphous
Handbbok. ICCROM, Rome. calcium carbonate into calcite in the sea urchin larval spicule. Proceedings of
Boynton, R.S., 1980. Chemistry and Technology of Lime and Limestone. John Wiley & the National Academy of Science (USA), 17362e17366.
Sons, New York. Ruppin, R., Englman, R., 1970. Optical phonons of small crystals. Reports on Progress
Bragg, W.L., 1914. The analysis of crystals by the X-ray spectrometer. Proceedings of in Physics 33 (2), 149e196.
the Royal Society of London 89 (613), 468e489. Sharon, I., Gilboa, A., Shalev, Y., 2009. Tel Dor 2005: Preliminary Report. In: Hada-
Canti, M.G., 2003. Aspects of the chemical and microscopic characteristics of plant shot Arkheologiyot e Excavations and Surveys in Israel, vol. 121.
ashes found in archaeological soils. Catena 54 (3), 339e361. Surovell, T.A., Stiner, M.C., 2001. Standardizing infrared measures of bone mineral
Chu, V., Regev, L., Weiner, S., Boaretto, E., 2008. Differentiating between anthro- crystallinity: an experimental approach. Journal of Archaeological Science 28,
pogenic calcite in plaster, ash and natural calcite using infrared spectroscopy: 633e642.
implications in archaeology. Journal of Archaeological Science 35, 905e911. Weiner, S., 2010. Microarchaeology. Beyond the Visual Archaeological Record.
Dollimore, D., 1987. The thermal decomposition of oxalates. A review. Thermochi- Cambridge University Press.
mica Acta 117, 331e363. Zolotoyabko, E., Caspi, E.N., Fieramosca, J.S., Von Dreele, R.B., Marin, F., Mor, G.,
Duyckaerts, G., 1959. The infra-red analysis of solid substances. Analyst 84, 201e214. Politi, Y., Addadi, L., Weiner, S., 2010. The differences between bond lengths
Falini, G., Fermani, S., Gazzano, M., Ripamonti, A., 1998. Structure and morphology of in biogenic and geological calcite. Crystal Growth and Design 10 (3),
synthetic magnesium calcite. Journal of Materials Chemistry 8 (4), 1061e1065. 1207e1214.