EPR dosimetry of biohydroxyapatite below liquid nitrogen temperature

Hydroxyapatite, major component of all organic solid tissues, can be used as a sensitive biodosimeter based on the electron paramagnetic resonance (EPR) spectroscopy. However, the dosimetric signal of biohydroxyapatite overlaps with the so-called parasitic signals due to the close g factor values and broadening of the resonance line at room temperature. Moreover, the unsaturated parasitic signals possess the intensity comparable to the dosimetric resonance. All of these significantly complicates the dose determination and limits applicability mainly to the cases of relatively large accumulated dose. The negligibly saturated dosimteric spectrum can be at least partially separated at the liquid nitrogen temperatures (LNT) due to the strong saturation and suppression of the parasitic resonance lines and the linewidth shortening as shown in the present work. Moreover, the advances in the modern EPR equipment in the last two decades resulted in high sensitivity and stability of the signals measured. These are the key parameters along with the computer simulations for the precise dosimetric spectrum separation and processing. This could lead to the higher accuracy of the LNT EPR method proposed in the present work. To test the approach, the stepwise dose calibration of biohydroxyapatite over the range 0.5 - 20 Gy was made. The corresponding dosimetric signal measured at 70 K exhibited the linear dose response. The results suggest the applicability of the LNT EPR method in the retrospective dosimetry.


Introduction
Hydroxyapatite (HAp) is the most important bioceramic material and the major component of all organic solid tissues appearing as the hexagonal symmetrical crystal system.It is commonly known to be used in many applications, for example in water purification (removal of heavy metals, defluorination), protein purification, polishing of monoclonal antibodies,, as catalyst in organic syntheses, or orthopedic and dental implant fabrication [1][2][3].Pure artificial hydroxyapatite is synthetized by chemical reactions at high temperature [4].It occurs naturally mostly as biohydroxyapatite (bHAp), sometimes referred to as carbonate-hydroxylapatite, which is formed by living organisms as a reinforcing material.The bHAp has the same crystallographic structure (hexagonal symmetry), but also contains impurities and substitutions, mainly carbonate ions present as an adsorbed phase or a lattice substituent of PO4 3-or OH - group.Water is present on the surface of enamel and as an absorbent in the enamel crystal lattice [5].The bHAp is produced by amenoblastic cells at room temperature.The biohydroxyapatite crystals are grouped in clusters bounded together by interprismatic biohydroxyapatite, with random orientations and without special symmetry, similarly to ceramic materials.Different growth conditions lead to different properties of the HAp and bHAp as known from materials research, i.e., the method of a material production can heavily influence the properties of the material [6].
From a dosimetry perspective, the bHAp has two interesting features.First, almost every living organism has some amount of this material incorporated into its body.Second, it is a very sensitive biodosimeter, having the ability to trap charge and store it for a very long time, even hundreds of thousands of years [7].Therefore, biohydroxyapatite dosimetry is used to determine the accumulated doses after exposure of a sample to the artificial radiation [8][9][10].In addition, this is used to determine the age of various archaeological artifacts [11].The principle is relatively simple.Ionizing radiation causes charge carrier excitation, and part is trapped by the charge traps in a lattice.The number of filled traps is proportional to the accumulated dose and represents a record of the irradiation history.There are two ways how to count the traps and determine the dose -by stimulated luminescence or EPR measurement, and a dose calibration, because both ways are relative measurements.In the case of EPR approach, the depletion of electron traps does not occur, which allows virtually unlimited number of measurements, higher sensitivity, accuracy, reproducibility, and low sample quantity requirements [8].Therefore, EPR dosimetry of dental enamel is sometimes referred to as the gold standard of retrospective dosimetry [12].For this reason, EPR dating is possible complementary method to solve some limitations of radiocarbon dating.
For some periods, age determination using 14 C inherently suffers with high uncertainty due to the plateau periods on the radiocarbon calibration curve or the periods of fluctuating 14 C abundance in the atmosphere [13].Another future problem for radiocarbon dating is the dilution of the natural isotopic composition of carbon by the burning of fossil fuels [14].According to a model simulation [15], it will start to be impossible to distinguish between contemporary and historical (year 1955 and older) samples in the next several years.This poses a serious problem especially for forensic applications of radiocarbon dating such as wildlife crime investigations [16].A combination of EPR and radiocarbon dating can be a way of resolving this ambiguity; using EPR dosimetry for relative resolution between the contemporary and historical samples, i.e. between the doses lower than 100 mGy and the higher ones.
However, there is a fundamental and unresolved issue of EPR biodosimetry -the dosimetric signal of biohydroxyapatite is strongly overlapped with the so-called parasitic signals.This uncertainty impair reliability and struggle a development of the most optimal measurement procedure and, therefore, slowed down wide application of EPR dosimetry in common practice [5].The logical way to solve this problem should be a cooling which in general leads to better spectral resolution [17], unfortunately it is well known, that the dosimetric signal of bHAp saturate at low temperatures.However, the degree of saturation could be reduced by the lowering of the microwave power radiation, although the signal intensity can be also significantly negatively affected.But EPR measurements can be repeated arbitrarily many times and accumulate signal strengths, and in this way detecting even very weak signals.Therefore, it can be assumed that a combination of parameters may exist, allowing to take advantage of low temperatures.
The original experiments in this direction were carried out decades ago and since then the sensitivity of EPR spectrometers has significantly increased.A successful attempt actually has been made to measure doses at low temperatures in a very special rapid-passage mode [18], but this experimental approach has produced no practical application.Therefore, the present work focuses on the standardsetup and low-temperature EPR measurements to determine the optimal conditions for the dating and to test the reliability of the such dose reconstruction.The Q-band and XRD measurements of contemporary and historical samples of bHAp also will be compared with HAp to study possible differences affecting the dosimetric properties of the samples.

Sample preparation
Three types of samples were obtained for the experiments.The first biohydroxyapatite specimen is an ancient tooth from archaeologists' collection dated by radiocarbon dating to 16th century.The second sample is a few months old wisdom tooth, a gift from our colleague, extracted during a dental procedure.The third sample is a commercially available form of pure hydroxyapatite purchased from Sigma-Aldrich.
Tooth samples (ancient and contemporary ones) were prepared for measurement using the procedure recommended by [5].The teeth were cleaned with water and then ethanol.A tooth root was mechanically broken off and its remnants were separated from the enamel by the chemical treatment in an ultrasonic water bath with 5M NaOH for 24h.The removal of all dentin was checked by microscopy and finalized by mechanical scraping of the softened dentin.Washed and dried samples were finely ground and sieving to a size of 0.5 mm.The chemical etching solution of 42% phosphoric acid was used, in three successive 30s applications, and after washed and well dried again.Samples of chemically pure hydroxyapatite were prepared just by fine crushing and sieving to the same size as biohydroxyapatite.
The irradiation of samples was performed at a water phantom (isolated from the water by a plastic bag) using calibrated 60 Co source.The samples were thus exposed to the same doses of irradiation.

Experimental Techniques
Electron paramagnetic resonance (EPR) spectra of X-band (9.4 GHz) measurements were obtained at room temperature (290 K) and just below liquid nitrogen temperature (70 K) using a commercial Bruker EMXplus spectrometer cooling by liquid Helium.The spectrometer sensitivity is about 10 12 spins/mT, while microwave power of 0.01-13 mW, modulation amplitude of 0.1-0.5 mT and time constant of 160 μs were used.EPR spectra of Q-band (34 GHz) measurements were obtained at room temperature using a commercial Bruker ELEXSYS E580 spectrometer.
The crystalline structure of powders was characterized using a powder X-ray diffractometer (Empyrean, Malvern Panalytical) with Cu K<α> radiation (λ = 1.54151Å, at U = 45 kV, I = 30 mA).The X-Ray diffraction patterns were measured in the range 2θ from 5 till 120 degrees with a step 0.026° and time per step 99.45 s.The characterization of powders was made in the Bragg-Brentano geometry using scanning line detector 1D and Ni filter.

Structural and Morphological Analysis
First, to verify the purity of HAp and to compare with bHAp the XRD patterns of the samples were measured.The results of the comparison are shown in figure 1 and table 1.The results showed that the hydroxyapatite sample has a hexagonal phase, space group P 63/m, space group number 176, density 3.15 g/cm3 (Hydroxylapatite, Reference code: 96-901-1094, Chemical formula: Ca10P6O26H2 References Structure: Sudarsanan, K., Young, R. A., Acta Crystallographica, Section B, 25, 1534 -1543, 1969).There are no traces of contamination from other elements or material phases.

X-band EPR
The EPR spectra of the biohydroxyapatite (tooth) sample were measured at room temperature (blue) and at 70 K (red).They are shown in figure 2. As can be seen there, while most of the signals in EPR spectrum overlap at room temperature, the dosimetric signal becomes separated at 70 K.To check whether the signal saturation (because of the elongated spin-lattice relaxation time) [17] is an unpassable obstacle the microwave power dependence of the signal intensity was measured.The result can be seen in figure 3.There the dosimetric signal of bHAp measured at the same temperature (70 K) but with variable microwave power is shown.The lowest power (0.01262 mW) resulted in the suppressed "parasitic" signal at around 3340 G whereas the intensity of the dosimetric signal seems to be still good enough for, i.e., the spectral resolution has been improved.The microwave power dependence of the dosimetric signal peak-to-peak intensity is shown in figure 4. The goal was to estimate the saturation level.Considering the fact that EPR intensity is directly proportional to the square root of the power [19] and the double increase of the microwave power value used in each step of measurements, the intensity ratio of each two steps beyond the saturation should be: Ii+1/ Ii = √2 (here i is ith step).From this follows that √2/(Ii+1/ Ii) = 1 without saturation.However, when the saturation is present, this relation is not fulfilled and the fraction of 1 appears as also indicated in figure 4 in the form of percentage.The closest value of 89% was found for the 0.01262 and 0.02524 mW.In general, considering relative low doses of irradiation needed in the EPR dating, this slight saturation cannot be an obstacle.To check whether it is so and the dosimetric signal at 70 K maintains a linear dose response, the calibrated doses (see Experimental) were imparted to the sample and the corresponding spectra are shown in figure 5 (left panel).
The tooth with high accumulated dose was the object of the investigation.It was measured before irradiation and then stepwise after the increased dose of irradiation in each step at the same temperature (70 K).The obtained spectra were then numerically processed: the background was subtracted and the intensity of the dosimetric component of the signal was calculated using double integration to obtain the absorption itself.It is directly proportional to the paramagnetic particles concentration [17].The dose dependence of the absorption is shown in figure 5 (right panel).It can be seen that despite slight saturation, the signal at 70 K was still linearly proportional to the dose.Therefore, one can conclude that the EPR dosimetry at 70 K seems to be possible.From the calibration curve (figure 5, right panel) it is possible to determine the value of the unknown "zero" dose, i.e. the dose before irradiation, by a simple calculation.This accumulation dose in the historical tooth was calculated to be 291 mGy (intercept a is equal to -871.47, the slope b is equal to 2999.78; from this, considering y = a + bx and y = 0, x = 871.47/2999.78= 0.291 Gy = 291 mGy, see figure 5).To accurately determine the age of the tooth, the dose rate at the site of the finding would be needed, but it was no longer possible to perform in situ.However, our primary goal was not to date this specimen accurately, rather to check the functionality of our concept.For this purpose, a very rough estimate can be made based on the available literature, as it shows that in most cases the local dose rate is relatively close to the average value of 1 mGy/year [11].This means that the age estimate determined by EPR at low temperature is of the order of magnitude of the actual age of the sample and the concept seems to work, although a reproduction and validation of these results on a larger number of samples with the known dose rate will be required in future.
Considering the direct proportionality between the EPR signal intensity and the paramagnetic particle concentration [17] and the linear response to the accumulated dose, the sensitivity of the method is restricted to the sensitivity of the EPR spectrometer used, i.e., 10 12 spins/mT.The upper detection limit is restricted to the number of specific sites serving as the charge trapping centers in the bHAp.

Q-band EPR
X-band resolution was not sufficient to investigate the differences between bHAp and HAp.Better spectral resolution can be reached at higher microwave frequency, therefore Q-band measurements were applied (see Experimental).However, the higher frequencies also lead to faster signal saturation in addition, therefore Q-band measurements proved incompatible with cooling and room temperature was used.The EPR spectra of the ancient tooth and the contemporary tooth were exposed to the same high dose of irradiation to increase the EPR signal intensity.The comparison of these experimental EPR spectra are shown in figure 6.As one can see, the spectra are very similar.The small differences originate from different amount of sample used in each experiment and, probably, different amount of organics not fully dissociated by acid (see Experimental).From this follows that, the possible residuals which might enter the enamel from the environment where the ancient tooth had been stored had no influence on the origin of the dosimetric signal.The signals are different, however, some similarities can be observed as indicated by vertical lines in figure 7. The component specific only for bHAp is also stressed.This is the sign of the multicomponent origin of the dosimetric signal.One part is the same as in the HAp whereas the other one indeed can be carbon-related originating from bHAp considering the chemical formulae of these samples as discussed in the Structural and Morphological Analysis (HAp: Ca10P6O26H2 and bHAp: Ca10.00P6.00O26.14H2.60C0.02).

Conclusions
The presented results demonstrate that EPR dosimetry is possible in standard operation at low temperature.It promises the potential for improved resolution that could allow simplified processing of EPR spectra and increased measurement sensitivity.Next, the Q-band comparison showed that the residuals from the environment where the ancient tooth is stored do not enter it to affect the dosimetric signal.On the other hand, the incorporated carbon-related impurities contribute or affecte to the dosimetric signal, the signal probably has multicomponent origin including the signal of similar origin as in HAp.

Figure 3 .
Figure 3. Power dependence of X-band EPR spektra of biohydroxyapatite at 70 K.

Figure 4 .
Figure 4. Peak to peak intensity of dosimetric signal as a function of microwave power measured at 70 K.Percentages represent an estimate of the level of the signal saturation calculated as a fraction of the intensity to the applied power between given points.

Figure 5 .
Figure 5. Dose calibration of biohydroxyapatite (tooth) at 70 K.The spectra are shown in the left panel whereas the dependence of the double integral intensity (absorption) is plotted in the right panel.

Figure 6 .
Figure 6.EPR spectra of the ancient and contemporary tooth measured in Q-band at room temperature.Vertical lines indicate similarities between the spectra.Vertical dashed line stresses component specific only for the ancient tooth.

Figure 7 .
Figure 7. EPR spectra of the HAp and bHAp measured in Q-band at room temperature.Vertical lines indicate similarities between the spectra.Vertical dashed line stresses component specific only for bHAp.

Table 1 .
Unit cell parameters and crystallites size of the samples.