S-S-dating & S-T-dating
NEW TECHNIQUES OF ABSOLUTE DATING
OF QUATERNARY SEDIMENTS

A.I.Shlukov, M.G.Usova, S.A.Shakhovets, L.T.Voskovskaya


The other parts of paper:
Demonstration
Discussion

Description of S-S & S-T-techniques

Introduction

In our previous paper (Shlukov et al. 1993) the results of methodical studies were expounded and concerned with TL dating of quaternary sediments in natural quartz. These studies proved us in incorrectness of several methods, generally used in TL dating. In particular we have abandoned the invalid procedure of natural dose curve reconstruction by artificial irradiation of studied material. We interpreted from the other point of view the sunlight UV influence on the minerals, led to bleaching of pre-genetic lightsum during sedimentation, and put the new notion of "theoretical zero of TL geochronometer". We proved experimentally the availability of radioactive fading (R-fading) of age information in quartz, which considerably stronger then temperature fading (T-fading), generally taking into account as the only one.

Also we would like to note here that all stated facts can equally take place in other minerals, being used in dating (feldspars, carbonates). Moreover as our conclusions generally apply to mechanism of accumulating the age information in minerals, they have to affect the other dating methods, based on palaeodosimetry. OSL and ESR techniques distinction of principle consists only in difference of instruments for deciphering the age information, which do not influence the mechanism of it's accumulating.

As alternative we suggested the approach, novel in principle and free from detected drawbacks, allowing to decipher age information by TL of natural quartz and called by us "S-S technique" (Shlukov, Shakhovets, 1986).Besides elimination of detected drawbacks this technique is characterised by high expressness. This allowed to provide the principally new statistic approach in age characteristics of quaternary sections. High density of picking the samples, as from section's top to bottom as in layers extension, allowed to execute the objective rejection of distorted results, taking place for example because of non-clarified problem of "zero-point" in quaternary TL chronometry.

In suggested technique, after it's testing by dating more than 2000 samples, we revealed some limitations. The regional restriction was found as the most considerable one. Technique's application is guaranteed only within common distributive province of the material, which forms the quaternary sediments. Such province for Russian plain proved to be Scandinavian distributive province. We failed to receive representative series of dates in South Ukraine because of considerable mixture of the material of Caucasus and Carpathians origin. Predominance of Urals material did not allow to get the results in Orenburg region.

Additional methodical studies, purposed to overcome the revealed limitations, led to creation of "S-T dating technique". It's first experimental testing with rather valid results has been done in quaternary samples from Novgorod region.

S-S dating technique

We use the basic formula for calculation of the age (Shlukov et. al., 1993):

(1)

where:
t - sample's age;
S - lightsum of natural peak of thermostimulated luminescence of quartz with temperature of maximum 300-310 C (hereinafter referred to as «chronometrical» or "t-peak");
- lightsum of t-peak saturation;
- lightsum of t-peak after sample's exposition in non-filter light of mercury lamp with 120 Wt power and at a distance 250 mm (so called "theoretical zero", /Shlukov et. al, 1993/);
E - natural dose rate of gamma-radiation in studied layer;
- constant, defining the dose sensitivity of t-peak.
(Remark: This formula is corollary of solution of first order kinetic equation, though the second order equation is more valid. We kept less true expression, used by us at earlier stage, for remaining unification, after we made certain of the value of sistematic deviation does not exceed 15%)

Among 5 values required for age calculation the values S, and E are that one of direct measurement; and is determined by indirect way.

Receiving the value made the main concept of S-S-technique. Our studies stated that within vast territory of Russian plain the lightsum of t-peak of early-middle-quaternary sediments are concentrated near some constant value, which within 15% mistake coincides with lightsum of deliberately saturated sediment of late Cretaceous (about 30 mln.years). It made us to think that: firstly, the quartz, being in composition of quaternary sediments at this vast territory, has the common source of origin - Scandinavian massive; secondly, the lightsum value of saturation is constant for all quartz of Scandinavian origin.

We would not discuss this phenomenon in this paper. We shall only note that, initially proposed as hypothesis, it has got a good corroboration after analysis of more than 2000 samples, and today it can be confirmed as stated fact. During testing we revealed the limits of S-S technique application, which almost coincided with Scandinavian distributive province.

At base of stated phenomenon we used the sample from Upper-Cretaceous sediments as our laboratory standard, defining value for substitution in calculating formula of age (1). This sample is picked from Upper-Cretaceous quartz sands in sand-pit of glass-factory near Litkarino-town (Moscow region). The content of quartz in light fraction is 100%, the content of light fraction is not less than 98%. Granulimetric content prevails in range of 0.1-0.25 mm. Laboratory index of this sample is "LtQ".

After that we determined four values of right part of formula (1), we could get the rest unknown value of dose sensitivity by the sample with known age. For this purpose we used radiocarbon age of plant fossil 34 thousand years, included in layer of loam, picked out for TL analysis in quarry "Monchalovo" near Rjev-town. Calculation by reverse the normal order of formula (1) gave the value:

Field sampling

In order to eliminate the unskilled operations we followed immovable principle - the direct participation of TL-dating specialists within all time of field sampling. Samples have been picked as from natural fresh rock walls as from drill core.

During works in rock walls we made their opening to achieve natural moisture. In prepared vertical walls we drilled the blast horizontal hole by handle instrument at the depth of about 1 meter. While sampling sandy sediments we gathered material directly from boring instrument and made the package at once in light impervious packing.

While sampling fine-grain material we gathered it directly from opened wall with further concentration by water sieving. We took simultaneously maximum precautions against direct sunlight exposition, making all procedures under water layer not less than 30 sm.

We measured gamma-dose rate in studied layer with the help of scintillation searching radiometer of CRP-68-01 or CRP-68-02-type, located in prepared blast drilled hole. Models of radiometers, used by us, have the pointer indicator. The procedure of measurement included fixation of 30 momentary readings with interval of 10 sec. Average value of E and root-mean-square error were calculated by received momentary readings and than were used for age determining and instrumental error of it's measurement.

While sampling material from drill cores we tried to make the core maximum clear from surface material, which contains, as the rule, many dirty material from other layers. Radiometric measurements had been done by well-radiometer CRP-68-02 at the depth of sampling.

Radiometric measurements present the special case when using drive pipes and clay mortar. Drive pipes bring additional errors by means of screen effect, and clay substance brings additional radioactivity. We eliminated their influence with the help of empirical corrections when executing the works with drive pipes and clay mortars.

Amount of the sample, required for qualitative analysis, depends upon content of quartz fraction with grain's size 0.1-0.25 mm. We are needed in guarantee for qualitative analysis that we shall get not less than 250 mg of pure quartz within mentioned size after all procedures.

Laboratory preparation of the samples

The ultimate end of laboratory preparation is separating from the sample monomineral quartz fraction with grain's size 0.1-0.25 mm. For this purpose we worked out the standard procedure, which allowed to receive phase purity of monomineral quartz extract not less than 98% for the samples of Scandinavian distributive province. Enrichment procedure included the following steps:
- removal of clay by it's stirring up in water;
- water sieving with separating 0.1-0.25 mm fraction;
- treatment with concentrated hydrochloric acid up to termination of gas emission;
- treatment with concentrated hydrofluoric acid during 40 minutes;
- separating the light fraction in bromoform.

Later we substituted the expensive procedure of separation in bromoform by separating the fractions in electromagnetic separator CIM-1. During such procedure non-electromagnetic component did not differ from light fraction, received in bromoform.

When treating the sample with concentrated HF the cleaning of impurities takes place, as well as removal the outer of the quartz grains, that released us from necessity of taking into account the influence of alpha radiation, remaining the debatable question up today.

So, with the help of such complex treatment we have got very pure monomineral extract with phase quartz purity about 98-99% for overwhelming majority of the samples from Scandinavian distributive province. The morphology of natural glow curve served as the main criterion for estimation of extract's purity. Pure quartz is characterised by natural glow curve with predominance of well-expressed t-peak, slightly complicated by small peaks-satellites in front and dying parts (curve 1 at fig. 1).

Fig. 1 The typical glow curves of pure (1) and impure (3) natural quartz. 2 - chronometrical peak ("t-peak"); 4 - low-temperature satellite; 5 - high-temperature satellite.
All glow curves are submitted hereinafter deducting red-light background.

We consider the high-temperature peak not to be the peak of thermostimulated luminescence in it's classic meaning. It's main feature, allowed us to make such conclusion, is non-sensitivity to UV. Lightsum of this peak is not large as the rule and we ignored it's influence on the measurements.

We consider the satellite in low-temperature part of t-peak to be connected with remained mineral impurities, hardly dissolving in HF. We suppose that the main impurities here are orthoclase. If the low-temperature satellite had considerable lightsum (curve 3 at fig. 1), the receiving shape of glow curve was the criterion of sample's sorting out for further analysis.

TL analysis

Analysis of thermostimulated luminescence had been performed at GEO-TL-1 apparatus, elaborated and produced by A.I.Shlukov. This apparatus allows to analyse the sample's aliquot up to 100 mg with possibility of choosing the heating rate in the following range: 1, 2, 5, 10, 20 C/sec. The heating linearity in dynamic range was not worse than 2%. Temperature range is 20 - 550 C. Measurement of temperature was fulfilled with the help of chromel-alumel thermocouple, brazed to the stainless heating plate by high temperature solder. The temperature calibration was executed in the melting points of number of salts, covering the whole temperature range. Electronic scheme of heating governing allows to stop the heating in the whole temperature range with transferring to isothermal regime with exposition in range of 1/8, 1/4, 1/2, 1, 2, 4, 8, 16 and min.

Thermostimulated luminescence in apparatus was registered by photo multiplier of type FEU-85 with Sb-Cs photocatode. Decreasing of "red-light background" was achieved with the help of light filter of saturated solution of CuSO with 1cm layer. Registration of glow curves was executed at double-co-ordinate millivoltmeter of type H-306 and H-307.

Routine analysis was executed in range 20 - 550 C with heating rate 10 C/sec. Standard aliquot for single heating was 25 mg. If the quantity of the material was insufficient for single heating, we used lesser quantity with further recalculation of lightsum to the value of 25 mg. The truth of proportional recalculation with the error not more than 5% was stated at the base of specially performed experiment. The minimum quantity for confident measurement presented about 5 mg, though we have executed sufficiently successful analysis by aliquots about 1mg.

We fixed the lightsum of t-peak by the ordinate of glow maximum. The truth of substitution of measuring the square of elementary peak by the ordinate of maximum with the error not more than 5% was stated by specialised experiments as well.

We defined the lightsum as average value by 5 independent heatings for excluding random large deviations, and then we made calculation of mean square deviation , used in calculation of age instrumental error, simultaneously by deviation of separate measurements from average value S. While performing such operation we recognised that 5 measurements were insufficient for reliable value , but we were compelled to do so because of limitation of material's amount. Comparison with qualitative measurement within 20 heatings showed that we got rather similar values of .

Fig.2 Samples glow curves after UV-bleaching
1 - pure quartz; 2 - one of the variants of impure sample;
yellow - "theoretical zero" of t-peak.

For defining the value we exposed the part of the sample, spread in Petri cup as monograin layer, by light of mercury lamp, as mentioned above. Then we received the values and so by the same way as when measuring natural lightsum.

We used glow curves morphology after UV exposition as additional criterion of sample's rejection. The typical glow curve of pure sample is submitted by curve 1, and impure sample - by curve 2 at fig.2.

Values and were measured by laboratory standard LtQ, which had been treated by all above procedures. Laboratory standard LtQ after UV treatment was heated as well for additional control. The typical picture of complete set of glow curves of the sample, using for age determining, is submitted at fig. 3.

Fig. 3 The typical complete set of glow curves for sample's age defining
- lightsum of laboratory standard LtQ;
S - natural sample's lightsum;
- sample's lightsum after UV-bleaching.

We used the set of the values S, , and E, received for each sample, for calculation of sample's age by formula (1), and together with the set of the values , , and - for calculation of age instrumental error .

Age instrumental error was determined with the help of well-known calculating formula of indirect measurement's error:
(2)
which has the following condition in concrete case:

(3)

where: .

If the lightsum S of the sample coincides with the lightsum of the standard within 1.5-fold error, i.e. , we consider these results as out limited one and calculate the age by formula:

(4)

S-T technique

S-T technique essence consists in measuring the temperature of t-peak maximum together with lightsum. The dependence of situation of luminescence maximum upon lightsum exists for thermo-phosphors, following the second kinetic order (Antonov-Romanovskyi, 1968). Maximum replaces to the low temperatures when lightsum increases (fig. 4). As follows from theory the extent of this replacement synonymously depends on saturation level . Thus, in cases when the value is not constant, in contrast to S-S-technique, simultaneous measuring the lightsum and temperature of luminescence maximum provides us with data about unknown saturation level.


Fig. 4 Replacement of temperature of luminescence maximum
with changing lightsum of quartz t-peak.

We derived the following formula for calculation of age in S-T technique:

(5)

where normalised dose sensitivity ( is defined with the help of the sample with known age by reverse the normal order of calculation as it was done in S-S technique when determining the dose sensitivity . The following formula for calculation of was derived:

(6)
where:

Sst, Tst - the lightsum and temperature accordingly of t-peak maximum of our laboratory standard LtQ; S & T - the same values for speciment;
= 1.7 eV - energetic depth of electron traps, responsible for thermostimulated luminescence of t-peak;
k - Boltsman constant.

The first executed testing of S-T technique within samples collection of Novgorod region, which had anomalous large lightsum of t-peak, found out rather successful results. However we can consider these results as prior one because the suggested technique is required in additional theoretical and experimental elaboration.

Acknowledgements

We would like to thank all investigators for their active assistance and support in sample collection and chronostratigraphic interpretation.

Also acknowledged is the skilful discussion of Dr.V.Mejdahl and Dr.H.Jungner.

References

1. A.I.Shlukov et al, Nuclear Instr. and Meth. in Phys. Res., 73 (1993), p.p. 373-381.
2. A.I.Shlukov and S.A.Shakhovets, Certificate of invention No 1250041, USSR, 1984.
3. A.I.Shlukov, Certificate of invention No 1550382, USSR, 1989.
4. V.V.Antonov-Romanovskyi, Kinetic of Photoluminescence of Crystallophosphorus, "Nauka", Moscow, 1966.