The composition of carbon-containing materials from 14C-dated pottery:
preliminary results
G.I. Zaytseva1, E. D. Skakovskii2, P. M. Dolukhanov 3, A. Shukurov3, N. N. Kovaliukh4,
V. V. Skripkin4, E. Glavatskaya4
1. Institute for the History of Material Culture, Russian Academy of Sciences, St Petersburg,
Russia, [email protected]
2. Institute of Physical Organic Chemistry, National Academy of Sciences of Belarus, Minsk,
Belarus, [email protected]
3. Newcastle University, Newcastle upon Tyne NE1 7RU, U.K.,
[email protected]
4. Institute of Environment Geochemistry of the National Academy of Sciences of Ukraine,
Kiev, Ukraine, [email protected]
ABSTRACT
The exact association of the organic material with archaeological context is especially important
in the 14C dating. Therefore, direct radiocarbon dating of organic matter embedded into an
archaeological artefact is an especially attractive option. Such a material can be the pottery itself.
Food residue on the pottery is often dated, but the residue is not always available. The clay mass
of the pottery contains carbon, often directly visible upon breaking of the pottery. It is, of course,
possible that the carbon embedded into the pottery matrix is of geological origin, and therefore
unsuitable for archaeological dating. However, the relative content of the carbon in the clay mass
of pottery can be as high as 2–5%, which excludes its geological origin.
New techniques for 14C dating of organic matter found within the pottery mass using the
conventional method has been developed in the Kiev Radiocarbon Laboratory. In order to justify
the reliability of this approach, one must demonstrate that this carbon is related to the
archaeological event associated with the production of the pottery, and also that reliable 14C dates
can be obtained for this material. These dating techniques have never been used before.
Here we discuss the origin of the carbon-containing material in the clay mass of the
pottery using the nuclear magnetic resonance method.
Key words: archaeology, chronology, radiocarbon, pottery, carbon-containing materials,
Nuclear Magnetic Resonance
INTRODUCTION
Wood, charcoal, bone and soil are most often used for radiocarbon dating. It is not unusual,
however, that archaeological sites do not yield any wood, charcoal, or any other organic material
suitable for the dating. This happens especially often in southern regions where environmental
conditions preclude the preservation of organic materials. In such cases, the bone collagen can
also be destroyed, making radiocarbon dating problematic. Furthermore, it is often difficult to
establish confidently the association of the material dated with the archaeological event of
interest; this can lead to discrepancies between the archaeological and radiocarbon dates.
A possible solution is the direct measurement of organic matter included into the
archaeological artefact. Such an artefact can be pottery. Food residue on the pottery can provide
evidence of the time when the pottery was used, but such a residue is not always preserved or
formed. Neolithic archaeological sites usually contain large amounts of pottery, which is usually
used for the archaeological attribution of the site. The age of a pottery assemblage is commonly
estimated as the radiocarbon age of organic matter samples obtained from objects which are
deemed to be synchronous with the pottery (hearths, seeds, wooden structures, etc.). As noted
above, the association of such objects with the pottery can be difficult to establish confidently.
Hence, direct measurements of the organic carbon derived from the pottery matrix have been
repeatedly attempted since the 1950s. Kohl and Quitta (1964) were among the first to obtain 14C
dates for the early Neolithic Linear-Band-Ceramics Culture (LBK) by a direct dating of the
pottery carbon. Bonsail et al. (2002) provide a useful review of the methods of pottery dating.
The origin of the organic matter is the main problem in the direct dating of pottery. In particular,
it has to be clarified what organic compounds contain the datable carbon. We present our results
on the identification of the carbon-containing materials within the pottery matrix using the
nuclear magnetic resonance (NMR).
METHODS
Nuclear magnetic resonance is widely used to identify organic compounds in a sample (Ionin
and Ershov, 1970; Ionin et al., 1983; Harris and Wasylishen, 2010). This method is based on the
detection of signals from nuclei placed in a strong magnetic field and irradiated with radiofrequency waves. The electronic environment of a nucleus induces the so-called chemical shift
which depends on the nuclear structure. Another relevant parameter is the spin-spin coupling
which is sensitive to the abundance of magnetic nuclei. The signal intensity carries information
on the number of similar nuclei. Finally, the signal’s structure provides information about the
interaction of magnetic nuclei with their environment. In this study was applied the NMR
spectroscopy of 1H and 13C magnetic nuclei. Modern NMR techniques relies on the analysis of an
accumulated response to strong radio-frequency irradiation pulses followed by Fourier transform
(FT NMR), which allows to obtain strong enough signal even for a very low abundance of the
nuclei studied. The 1H and 13C NMR spectra can reveal the esters fatty acids, alcohols and other
organic compounds. Since the method is sensitive to the presence of double bonds in organic
compounds, carbon of geological origin it can be confidently identified.
We used the Bruker AVANCE 500 NMR spectrometer (with the working frequency of
500 MHz for 1H nuclei and 126 MHz, for 13C nuclei) to analyse solutions extracted from both
food residue on, and the charcoal within Neolithic pottery samples. A black, charcoal-like layer
can easily be seen within freshly broken pottery fragments (Fig.2). The chemical composition of
this organic matter and that of the food residue were analysed separately. The pottery fragments
and the food residue were powdered, and the organic compounds were extracted by dissolving
the powder in hexane. Hexane was then removed and the residue was dissolved in deuterated
chloroform (CDCl3). The solution was then NMR analysed for 1H and 13C nuclei.
NMR spectral lines were identified by comparison with reference compounds spectra,
including olive oil for the lines of triglycerides, cetyl palmitate СН3(СН2)14С(О)О(СН2)15СН3
and decyloleat СН3(СН2)9ОС(О)(СН2)7СН=СН(СН2)7СН3 for esters; octyl ether [CH3(CH2)7]2O
for ethers; 1-octadecanol [СН3(СН2)]17ОН for fatty alcohol signals; stearic acid
[СН3(СН2)]16СООН as a fatty acids representative; octadecan (paraffin СН3(СН2)16СН3); and
beeswax. To record the spectra, these compounds were also dissolved in CDCl3.
The relative content of organic compounds in the samples can be determined from the intensity
of the spectral lines arising from double bounds (I-the chemical shift δ = 5.35), as well as those
from the triglycerides of unsaturated fatty acids (II- δСНО = 5.30, δСН2О = 4.32 and 4.15,), esters
(III - δСН2ОС(О) = 4.06), alcohols (IV- δСН2О = 3.64), fatty acids (V- δСН2С(О)ОН = 2.25÷2.40) and
ethers (δСН2О = 3.38; only traces of these can usually be found). In addition, various aromatic
compounds were found in the solutions (VI-VIII- δСН =6.8–7.8).(Fig.3,4) Intensive signals
around.6 – 2.2 ppm are related to protons of the aliphatic chains of compounds mentioned above.
RESULTS AND DISCUSSION
Methods of 14C dating of carbon-containing materials from pottery have recently been developed
in the radiocarbon laboratory in Kiev using the conventional technique (Skripkin and Kovaliukh,
1998; Zaytseva et al., 2009). Pottery is one of the most suitable materials for dating because of
its close association with archaeological events. It plays an especially important role in the
absolute chronology of the Neolithic in the Southern regions of Russia, including Northern
Caucasus and the Caspian Sea region. Organic materials, including wood, bone, etc., are poorly
preserved there due to the arid and hot climate. Pottery is the main dating materials in such cases.
About 200 new radiocarbon dates for the Neolithic in the Southern and the Middle Volga regions
of Russia have already been produced by the direct dating of pottery using these methods
(Vybornov, 2008; Zaytseva et al., 2009).
The origin of the carbon-containing materials in the pottery can be problematic, and it is
important to verify if those materials are directly related to the archeological context. The
situation with the food residue is much simpler as this material is certain to reflect the time when
the pottery was in use. However, such a residue is rather rare, in part because it often used to be
removed during the clean-up of the samples, e.g., to reveal ornamental patterns. This makes the
direct dating of the pottery an especially attractive alternative. However, the carbon-containing
materials within the clay mass of the pottery can be of geological origin, unrelated to the any
archeological events. Here we compare the 14C dates from the pottery itself, to those from the
food residue on it, as well as from charcoal, bone remains and wood. For this purpose, we used
pottery samples collected from the Neolithic sites of the Russian regions shown in Fig. 1
Numerous 14C dates from different materials are available for the Rakushechny Yar site
which belongs to the early Neolithic. This site, located in the Rostov Region (Fig. 1), has been
extensively excavated by Belanovskaya (1995). New, radiocarbon dates for this site are
presented in Table 1. Calibration of radiocarbon ages was carried out using OxCal 4.1 with the
IntCal 09 calibration curve (Bronk Ramsey, 2009). The range of the calibrated age given in the
last column of the Tables 1 and 2 corresponds to the 95% confidence interval.
Table 1.
14
C dates obtained for various materials from the Rakushechny Yar Neolithic site
Lab. index
Material
Ki-6476
Ki-6477
Ki-6475
Ki-6480
Ki-6478
Ki-6479
Ki-11090
Ki-11091
Le-5344
IGSD-1371
IGSD-1371
food residue, layer 20
food residue, layer 20
food residue, layer 20
food residue, layer 15
food residue, layer 15
food residue, layer 15
pottery carbon, layer 12
pottery carbon, layer 11
Shell, layer 15
Viviparous shell
Unio shell
14
C Age
(BP)
6930±140
7860±130
7690±110
7040±100
6930±100
6825±100
7090±110
6955±160
7180±250
7685±70
8050±70
Calibrated Age
(Cal BC, 2σ)
6065–5559
7063–6457
6901–6260
6076–5722
5991–5638
5971–5556
6203–5728
6160–5557
6479–5617
6645–6419
7297–6697
It is clear from Table 1 that the dates produced from the food residue and from the claymass carbon agree well within errors. On the other hand, radiocarbon dates from shells are older.
This can be attributed to the reservoir effect.
In order to quantify the reservoir effect, we performed the following comparisons. The
main sources of carbon in pottery are peat, shells, plant remains and other materials used as
temper to improve the technological characteristics of the pottery (Vasilieva, 1999). To test the
reliability of the direct dating of these materials embedded in the pottery matrix, microscopic
shell pieces (typical size of a few microns) were manually extracted from the pottery matrix of a
sample from the Tektensor site in the Caspian Sea region (Table 2). A fragment of pottery from
the Tektensor site is shown in Fig.2. The shell inclusions were AMS dated at the radiocarbon
laboratory of Uppsala University (Sweden). The remaining pottery mass was treated with
hydrofluoric acid to eliminate any carbonates. The carbon remaining was then also AMS dated in
the same laboratory. Another parameter measured was the stable isotope 13C content, given
below in terms of δ13C = δ13C/ δ12C.
The age of the shell inclusions was determined as 7235 ± 45 BP with δ13C= −13.3‰ (Ua35226). The age of the carbon remaining after the acid treatment is 6695 ± 40 BP, with δ13C=
−27.7‰. The pottery fragments were further separated into an inner layer (‘the ‘terracotta’, redcoloured portion) and the outer layers (of a darker colour). The ‘terracotta’ layer is free of
carbon. The 14C age obtained for the carbon-containing part was determined as 6695±40 BP,
with δ13C= −27.7‰ (Ua-35227). This age is in perfect agreement with that obtained at the Kiev
Laboratory with the liquid scintillation techniques: 6630±80 BP (Ki-14137).
Table 2. 14C dates from different fragments of the clay mass of pottery from the Tektensor
Neolithic site, Caspian Sea region (Astrakhan district).
Lab. Index
Material
Ua-35226
Ua-35227
Ki-14137
shells from pottery
charcoal pottery
pottery, entire sample
14
C Age
(BP)
7235±45
6695±40
6630±80
δ 13C
( ‰, PDB)
−13.3
−27.7
–
Calibrated Age
(Cal BC, 2σ)
6220–6000
5670–5520
5730–5460
It remains to determine which carbon-containing compounds are contained in the food
residue and in the clay mass of the pottery. The following carbon-containing compounds can be
present in the clay mass:
1. Carbon of the organic matter in the clay, which has a geological origin; its age can be
reasonably expected to be much older than that of the pottery.
2. Carbon derived from temper (grass, straw, chaff, dung, ground shells, etc.); its age should
be the same as that of the manufacture of the pottery. However, the reservoir effect has to
be allowed for in the case of shells.
3. Carbon absorbed from fuel during the pottery firing in the kiln.
4. Carbon absorbed from the surrounding soil.
5.
To separate these contributions, we used the NMR methods. Suitable samples originate from
the Zamost’e Neolithic site in the Central Part of European Russian Plain, Moscow Region,
shown in Fig. 1, which is located in the forest zone and so the finds represent wood, charcoal,
peat, as well as pottery with food residue. 14C dates for this site, taken from Timofeev et al.
(2003), are presented in Table 3.
Table 3
14
C dates of the Zamost’e Neolithic site (Moscow Region , Fig.1)
Lab. index
Material
GIN-6198
GIN-6557
GIN-6564
GIN-7985
peat
peat
peat
peat
14
C age, BP
6680±100
6850±60
7050±40
6290±40
Calibrated
age, cal BC
5663–5485
5786–5641
5986–5843
5317–5132
Calibrated
age, cal BC
5746–5469
5840–5634
5994–5809
5361–5080
GIN-7988
GIN-7986
Ki-15031
Ki-15030
Ki-15032
peat
wood
charcoal
food residue
food residue
7200±90
7000±70
6730±120
6440±120
6300±130
6202–5928
5980–5801
5727–5530
5526–5298
5465–5069
6231–5885
5991–5730
5842–5470
5621–5079
5482–4936
The 14C dates from the food residue are somewhat younger than those from peat and
wood. Unfortunately the pottery carbon was not dated, but our experience with the Rakushechny
Yar site suggests that the 14C ages of the food residue and the pottery carbon should be consistent
with each other (Table 2).
The NMR spectra of the food residue and carbon-containing material of the pottery clay
mass for the Zamost’e samples are presented in Fig. 3 and 4.
The 1Н spectra (Fig. 3) reveal the presence of similar, rather complex chemical
compounds in both materials. Among such compounds are those with double bonds (C=C),
ethers, alcohols and fatty acids, as well as benzene derivatives. These compounds are more
abundant in the food residue than in the pottery itself: the total organic mass content in the food
residue is 40 times larger than in the pottery clay mass. The 1Н NMR spectra indicate that the
food residue contains other complex compounds which are as yet difficult to identify. We have
been able to indentify following organic compounds:
1. Compounds with double bounds (II), having chemical shifts δСНО = 5.30, δСН2О = 4.32
and 4.15 ppm.
2. Esters (III), having chemical shifts δСН2ОС(О) = 4.30 and 4.06 ppm.
3. Alcohols (IV), with chemical shift δСН2О = 3.64 ppm.
4. Fatty acids (V ), chemical shift δСН2С(О)ОН = 2.25÷2.40 ppm and ethers, chemical shifts
δСН2О = 3.38 ppm.
5. Various aromatic compounds, with chemical shifts δСН =6.8–7.8 ppm (VI–VIII).
The spectra of the food residue and the clay mass of the pottery were carefully compared and
interpreted (see Figs 3 and 4). The high peaks around 0.6–2.2 ppm are related to the protons of
the aliphatic chains of the compounds mentioned above.
The pottery mass has a relatively high relative content of ethers (δ = 4.05 and 4.3 ppm)
which can originate in vegetable waxes.
The 13C spectra of the food residue and pottery are visibly different (Fig. 4). While the
pottery spectra accumulated after 10 hours of measurement show only weak signals of the
residual hexane and paraffin, the spectra of the food residue contain more lines, including those
similar to the unsaturated (double-bond) compounds in the 1H spectra.
We conclude that the organic-containing materials in the food residue and pottery itself
are likely to have the same origin. One can suggest that the organic material in the pottery matrix
may have penetrated from the inner pot surface during food cooking.
These are the first results of this kind, and they need to be confirmed by further detailed
and careful studies. If confirmed, they strongly suggest that the 14C dates obtained from the
pottery itself can be considered as a reliable indicator of the time when the pottery was in use.
CONCLUSIONS
We have studied the origin of the organic-containing materials in the pottery using Nuclear
Magnetic Resonance techniques; to our knowledge, this is the first attempt of this kind. The
results obtained demonstrate that the food residue and the pottery matrix contain practically the
same organic compounds, even if the relative abundances of various compounds are different in
the two materials. It appears quite plausible that the organic compounds in the pottery matrix and
in the food residue have the same origin, i.e., they arise from the food cooked in the pots. This
implies, in turn, that the organic matter of the pottery temper was removed from the pottery
matrix at an earlier stage, e.g., as CO2 during the firing. If further studies will have confirmed
these suggestions, we can conclude that the 14C dates obtained from the carbon in the pottery
matrix can be used to date the period when the pottery was in use, alongside the food residue.
We plan to study the pottery from other sites in the future.
This research is supported by the FP6 project NEST No. 028192 FEPRE and the grant
.RFFR No. 10-06-00096а
REFERENCES
Belanovskaya, T.D., 1995. Iz drevneishego proshlogo Nizhnego Podon’ya. Poselenie vremeni
neolita i eneolita Rakushechnyi Yar. St.Petersburg, St. Petersburg University Press (in
Russian).
Bonsail, C., G. Cook, J. L. Manson and D. Saderson, 2002. Direct dating of Neolithic pottery:
progress and prospects. Documenta Praehistorica , XXIX, 47–59.
Bronk Ramsey, C., 2009. Bayesian analysis of radiocarbon dates. Radiocarbon, 51, 337–360.
Dolukhanov, P. M, Shukurov, A., Davison, K., Sarson, G., Gerasimenko, N. P.,
Pashkevich, G. A., Vybornov, A., Kovaliukh, N. N., Skripkin, V. V., Zaitseva, G. I. and
Sapelko T. V., 2009. The spread of the Neolithic in the South East European Plain:
radiocarbon chronology, subsistence and environment. Radiocarbon, 51, 783–795.
Ionin B.I., Ershov B.A.,1970. NMR spectroscopy in organic Chemistry. New York: Plenum
Publising Corporation. 195 p.
Ionin B.I., Ershov B.A., Kol’tsov A.I. 1983 NMR-spectroscopy in the organic chemistry.
Leningrad. Chimiya press. 272 p. (in Russian)
. Harris R. K and. Wasylishen R (editors), 2010. Encyclopedia of Magnetic Resonance, John
Wiley: Chichester. DOI: 10.1002/9780470034590.emrstm0211.
Skripkin, V. V. and Kovalyukh, N. N., 1998. Recent developments in the procedures used at the
SSCER Laboratory for the routine preparation of Lithium Carbide. Radiocarbon, 40,
211–214.
Timofeev V.I., Zaitseva G.I., Dolukhanov P.M., Shukurov A.M.2004159 p. (in Russian).
Radiocarbon chronology of the Neolithic of Eurasia. Teza. St.-Petersburg
Vsilieva I.N. 1999. Fictility of the population of the Northern Caspian Sea region in the
Neolithic. The Questions of Archaeology of the Volga river regions. Samara. Issue 1. Pp.
77-89 (in Russian)
Vybornov, A. A., 2008. The Neolithic of the Volga–Kama region. Samara University Press. 488
pp. (in Russian).
Zaitseva, G., Skripkin, V, Kovaliukh, N., Possnert, G., Dolukhanov, P. and Vybornov, A., 2009.
Radiocarbon dating of Neolithic pottery. Radiocarbon, 51, 795–803.
FIGURES CAPTIONS
Fig.1. The Neolithic sites of the Southern and the Low Volga river basin investigated:
(1) Racushechny Yar, (2) Matveev Kurgan, (3) Kairshak I–III, (4) Kugat IV,
(5) Tektensor I, III, (6) Burovaya, (7) Varfolomeevskaya, (8) Lebyazhinka,
(9) the Yelshanian group of sites, (10) Chekalino IV, (11) Maksimovka, (12) Ivanovskoe,
(13) Dzangar, (14) Il’inka, (15) Zamost’e.
Fig.2. The profile of the pottery fragment from the Tektensor Neolithic site.
Fig.3. The 1H NMR spectra of the components of the food residue and pottery of the Zamost’e
Neolithic site. I: compounds with double bounds; II: triglycerides of fatty acid esters;
III: ethers; IV: alcohols; V: fatty acids and ethers; VI-VIII: various aromatic compounds.
Fig. 4. The 13C NMR spectra of the components of the food residue and pottery of the Zamost’e
Neolithic site. I: compounds with double bounds; II: triglycerides of fatty acids esters;
III: ethers; IV: alcohols; V: fatty acids and ethers (δСН2О = 3.38); VI-VIII: various
aromatic compounds.
Food residue
V
SSolvent
IV
I II
VI
VI
II
7.0
6.0
5.0
II III
4.0
3.0
2.0
1.0
2.0
1.0
ppm (t1)
Pottery itself
Solvent
V
I II
VI
IV
VI
7.0
ppm (t1)
III II III
6.0
5.0
4.0
3.0