ENDOR study of irradiated tooth enamel

γ- and x-irradiated tooth enamel has been studied by EPR and ENDOR. Radiation-induced EPR spectrum of tooth enamel was found to be a superposition of signals with dominant contribution determined by CO₂- radicals. Two types of these radicals were observed: ordered and disordered centers. EPR spectra...

Повний опис

Збережено в:
Бібліографічні деталі
Опубліковано в: :Semiconductor Physics Quantum Electronics & Optoelectronics
Дата:1999
Автори: Ishchenko, S., Vorona, I., Okulov, S.
Формат: Стаття
Мова:English
Опубліковано: Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України 1999
Онлайн доступ:https://nasplib.isofts.kiev.ua/handle/123456789/117926
Теги: Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
Назва журналу:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Цитувати:ENDOR study of irradiated tooth enamel / S. Ishchenko, I. Vorona, S. Okulov // Semiconductor Physics Quantum Electronics & Optoelectronics. — 1999. — Т. 2, № 1. — С. 84-92. — Бібліогр.: 29 назв. — англ.

Репозитарії

Digital Library of Periodicals of National Academy of Sciences of Ukraine
id nasplib_isofts_kiev_ua-123456789-117926
record_format dspace
spelling Ishchenko, S.
Vorona, I.
Okulov, S.
2017-05-27T16:03:18Z
2017-05-27T16:03:18Z
1999
ENDOR study of irradiated tooth enamel / S. Ishchenko, I. Vorona, S. Okulov // Semiconductor Physics Quantum Electronics & Optoelectronics. — 1999. — Т. 2, № 1. — С. 84-92. — Бібліогр.: 29 назв. — англ.
1560-8034
PACS 61.72; 76.30; 76.70
https://nasplib.isofts.kiev.ua/handle/123456789/117926
γ- and x-irradiated tooth enamel has been studied by EPR and ENDOR. Radiation-induced EPR spectrum of tooth enamel was found to be a superposition of signals with dominant contribution determined by CO₂- radicals. Two types of these radicals were observed: ordered and disordered centers. EPR spectra of both CO₂- centers are described by axial g-tensor with g = 1.9975 and g ⊥= 2.0021 with g || c. The ENDOR spectrum of unannealed enamel powder consists of a singlet at Larmor frequency of ³¹P nuclei and doublet at Larmor frequency of ¹H nuclei. Samples annealing at T = 200-250 ⁰C resulted in the destruction of disordered centers and appearence of superhyperfine structure of ENDOR spectra. Its analysis with advanced the powder ENDOR theory allows to determine the superhyperfine constants and to find for the first time that the ordered CO₂- radical is located in B sites (phosphorous position) of bioapatite lattice. This substitution is accompanied by the shift of the nuclei of the first ³¹P shell towards the defects by 0.04 nm and the formation of the OH vacancy in the nearest radical surroundings.
The authors wish to thanks Prof. A.B. Roitsin for consultations on the theory of powder ENDOR and Dr. S.V. Virko for assistance in computer calculations.
en
Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України
Semiconductor Physics Quantum Electronics & Optoelectronics
ENDOR study of irradiated tooth enamel
Article
published earlier
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
collection DSpace DC
title ENDOR study of irradiated tooth enamel
spellingShingle ENDOR study of irradiated tooth enamel
Ishchenko, S.
Vorona, I.
Okulov, S.
title_short ENDOR study of irradiated tooth enamel
title_full ENDOR study of irradiated tooth enamel
title_fullStr ENDOR study of irradiated tooth enamel
title_full_unstemmed ENDOR study of irradiated tooth enamel
title_sort endor study of irradiated tooth enamel
author Ishchenko, S.
Vorona, I.
Okulov, S.
author_facet Ishchenko, S.
Vorona, I.
Okulov, S.
publishDate 1999
language English
container_title Semiconductor Physics Quantum Electronics & Optoelectronics
publisher Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України
format Article
description γ- and x-irradiated tooth enamel has been studied by EPR and ENDOR. Radiation-induced EPR spectrum of tooth enamel was found to be a superposition of signals with dominant contribution determined by CO₂- radicals. Two types of these radicals were observed: ordered and disordered centers. EPR spectra of both CO₂- centers are described by axial g-tensor with g = 1.9975 and g ⊥= 2.0021 with g || c. The ENDOR spectrum of unannealed enamel powder consists of a singlet at Larmor frequency of ³¹P nuclei and doublet at Larmor frequency of ¹H nuclei. Samples annealing at T = 200-250 ⁰C resulted in the destruction of disordered centers and appearence of superhyperfine structure of ENDOR spectra. Its analysis with advanced the powder ENDOR theory allows to determine the superhyperfine constants and to find for the first time that the ordered CO₂- radical is located in B sites (phosphorous position) of bioapatite lattice. This substitution is accompanied by the shift of the nuclei of the first ³¹P shell towards the defects by 0.04 nm and the formation of the OH vacancy in the nearest radical surroundings.
issn 1560-8034
url https://nasplib.isofts.kiev.ua/handle/123456789/117926
citation_txt ENDOR study of irradiated tooth enamel / S. Ishchenko, I. Vorona, S. Okulov // Semiconductor Physics Quantum Electronics & Optoelectronics. — 1999. — Т. 2, № 1. — С. 84-92. — Бібліогр.: 29 назв. — англ.
work_keys_str_mv AT ishchenkos endorstudyofirradiatedtoothenamel
AT voronai endorstudyofirradiatedtoothenamel
AT okulovs endorstudyofirradiatedtoothenamel
first_indexed 2025-11-25T23:07:37Z
last_indexed 2025-11-25T23:07:37Z
_version_ 1850581028734763008
fulltext 8 4 © 1999, Institute of Semiconductor Physics, National Academy of Sciences of Ukraine Semiconductor Physics, Quantum Electronics & Optoelectronics. 1999. V. 2, N 1. P. 84-92. PACS 61.72; 76.30; 76.70 ENDOR study of irradiated tooth enamel S. Ishchenko, I. Vorona, S. Okulov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine, Kyiv, 252028, Ukraine Abstract. γ- and x-irradiated tooth enamel has been studied by EPR and ENDOR. Radiation- induced EPR spectrum of tooth enamel was found to be a superposition of signals with dominant contribution determined by CO 2 - radicals. Two types of these radicals were observed: ordered and disordered centers. EPR spectra of both CO 2 - centers are described by axial g -tensor with ||g = 1.9975 and g ⊥= 2.0021 with cg |||| . The ENDOR spectrum of unannealed enamel powder consists of a singlet at Larmor frequency of 31P nuclei and doublet at Larmor frequency of 1H nuclei. Samples annealing at T = 200-250 0C resulted in the destruction of disordered centers and appearence of superhyperfine structure of ENDOR spectra. Its analysis with advanced the pow- der ENDOR theory allows to determine the superhyperfine constants and to find for the first time that the ordered CO 2 - radical is located in B sites (phosphorous position) of bioapatite lattice. This substitution is accompanied by the shift of the nuclei of the first 31P shell towards the defects by 0.04 nm and the formation of the OH vacancy in the nearest radical surroundings. Keywords: tooth enamel, electron-nuclear double resonance, radiation-induced electron-paramag- netic resonance. Paper received 11.12.98; revised manuscript received 01.04.99; accepted for publication 19.04.99. 1. Introduction The investigation of bioapatites has been started a few decades ago. Late in the 60s apatites have been consid- ered as a perspective laser material and studied by dif- ferent techniques. Radiospectroscopic investigations have been directed to clarification of mechanisms of in- corporation of impurity atoms such as carbon, silicon, sulphur and different cations in apatites lattice. The com- plex study of naturae, synthetic and bioapatites has re- sulted in building up of laser operating in IR region [ 1]. Besides, calcium apatites had attracted considerable at- tention because they form a base of the important bio- logical tissues such as bone and dental enamel. In these tissues crystallites of hydroxyapatite are organized into small prisms plunged into organic substance. Both prisms and crystallites reveal certain ordering. The composition and orientation of crystallites are believed to be closely related to the total state of organism�s health. Therefore, the information about their changes can be used in med- icine for early diagnostics of some diseases [2]. Besides, ionizing irradiation of bioapatites induces paramagnet- ic centers, the number of which is proportional to the irradiation dose. This phenomenon is used in retrospec- tive dosimetry to determine radiation dose received by people and animals [3-5] and in EPR dating of archeo- logical findings [6, 7]. Tooth enamel is the most interesting calcified bio- logical tissue. It contains a mineral phase representing about 94-98 % of the total weight, the rest is formed by water and organic matter. The EPR signal near g = 2.0 appeared after irradiation by γ- and x-rays is well known [8-14]. However, the nature of this radiation-induced sig- nal is not clarified completely in spite of many publica- tions devoted to its study. EPR spectrum in the irradiated tooth enamel has been found to be a composite signal [13]. It leads to variation of the line shape depending on experimental conditions and sample history. This is the reason of discrepancies among the data reported in different papers and leading to uncertainty and mistakes in EPR dosimetry. The CO 2 -, CO 3 3-, CO 3 -, CO- radicals [13], O- center [14] S. Ishchenko et al.: ENDOR study of irradiated tooth enamel 85SQO, 2(1), 1999 and the so-called �background signal� with g = 2.0045, ∆H pp = 0.64 mT give the main contribution to the mentioned radiation-induced EPR spectrum. The back- ground signal is observed in the enamel samples before irradiation and dominates at irradiation doses up to 1 Gy. Note that radicals can be located in different sites of hydroxyapatite lattice resulting in the increase of the number of the EPR components. Besides, the irradiation of tooth enamel induces short-half-life radicals. Some centers are destroyed by temperature annealing whereas some new defects can be formed in accordance with annealing processing. However, the contributions of the mentioned centers are not equivalent. A dominant contribution is caused by the centers of one type. Some authors have assigned this more intensive signal to CO 3 3- radical [8-10, 15] while others have classified it as CO 2 - radical. The most convincing evidence of CO 2 - interpretation is the pres- ence of g-tensor component equal to 1.997. This value is characteristic for CO 2 - and varies slightly depending on radical surroundings. Besides, the recent study of CO 3 3- in synthetic hydroxyapatite [16] allows to determine the following g values of the radical: g x = 2.0045, g y = =2.0034, g z = 2.0014. These parameters do not describe the dominant EPR signal in tooth enamel. Callens et al. [17] have carried out the EPR study of intact and powdered human tooth enamel both unheated and dried at 400 0C. Existence of two main EPR components with similar parameters has been found. One of them does not exhibit anisotropy and determined by disordered centers. This component has been attributed to CO 2 - radical in organic matrix of enamel and/or sur- face CO 2 - radical. Another EPR component has been assigned to a bulk CO 2 - radical. It has been noted that the annealing results in the destruction of the disordered centers. However, the position in the lattice has not been determined for both radicals. ENDOR is the most useful technique to define the location of a paramagnetic center in crystalline lattice. A few ENDOR measurements were carried out on tooth enamel [8, 10, 18, 19]. The dominant ENDOR signals observed in these experiments were the structureless sin- glet line at Larmor frequency of 31P nuclei and singlet or doublet line at Larmor frequency of 1H nuclei. Such ENDOR spectra do not allow to obtain information about the location of CO 2 - radical in hydroxyapatite lat- tice. More informative ENDOR results have been ob- tained on the synthetic hydroxyapatite. The superhyper- fine structure of the 31P and 1H ENDOR spectra per- mits to determine the location of CO 3 3-[16] and O-[14] radicals in this material. This article is devoted to the detailed ENDOR study of irradiated tooth enamel to define the location of CO 2 - in hydroxyapatite crystallites. The structure of 1H and 31P ENDOR lines was observed on enamel samples dried at 250 0C. Its analysis allowed to obtain the information about CO 2 - surroundings and to determine the radical location in bioapatite lattice. 2. Materials and methods The powder samples were prepared according to the tra- ditional procedure of retrospective EPR dosimetry [3]. The paramagnetic centers were created by irradiation with x- or γ- rays. The absorption dose was estimated to be approximately 10 kGy. The heating of the samples was carried out on air at temperatures 250-300 0C dur- ing 30-60 minutes. Such annealing conditions were opti- mum to observe the superhyperfine structure of ENDOR spectra. ENDOR spectra were recorded using EYa-1301 su- perheterodyne spectrometer operating in the 3-cm range of wavelengths. Measurements were performed at the temperatures 77, 4.2 and 1.5 K. A sample was placed along the axis of a TE 011 cylindrical microwave cavity. A radiofrequency (rf) field inducing nuclear transitions was generated by four rods driven across the cavity parallel to its axis and connected in accordance with Helmholtz scheme. The amplitudes of rf field were 0.01-0.05 mT. The ENDOR signals were registrated by pulse modula- tion at 1.5 kHz of the applied rf power and phase-sensi- tive detection. The ENDOR spectra were recorded for powder samples of the human and swine tooth enamel. On heated samples the structure of ENDOR lines was obtained with microwave power of approximately 5 µW at liquid helium temperature. The dependence of ENDOR spectra on the type of irradiation (x- or γ- rays) and type of enamel (human or swine) was not observed. 3. The theory of powder endor The orientation dependencies of ENDOR spectra can be studied on powders if g -tensor of paramagnetic cen- ter is anisotropic. During ENDOR measurements the magnetic field H is fixed in certain position of a powder EPR spectrum. This induces the resonance only for those centers which have definite orientation relative to the direction of the magnetic field. The changing of H value permits to observe the centers with different orientations separately, i.e., to registrate the angular variation of ENDOR lines. The obtained dependencies are similar to that of single crystals if superhyperfine A-tensor and elec- tron g -tensor have axial symmetry and coinciding di- rections. In general case, ENDOR spectrum is partly averaged. However, the lines of such spectrum alter their spectral positions while changing the value of fixed mag- netic field. This permits to obtain the information about the parameters of superhyperfine interaction. The above technique has been named �orientation ENDOR selec- tion� [20, 21]. The general theory of powder ENDOR is very complicated. Its particular cases have been described before [21-24]. It is appropriate to present the simplified expressions which were used in this work. The expres- sions have been obtained in the first order of the pertur- bation theory and given the satisfactory description of our experimental results. Note that these expressions do S. Ishchenko et al.: ENDOR study of irradiated tooth enamel 8 6 SQO, 2(1), 1999 not consider the effect of alternating fields, relaxation processes and transition probabilities. They contain also some limitations which will be pointed out below. The spectrum of powder ENDOR caused by unpaired electron and nucleus with spins S = I = 1/2 can be described in g -tensor axes frame (see Fig. 1) by the following expression: ∫ ∫ ∑ ×−= 2/ 0 2 0 0 )),((sin),( π π ϕθθυ m mHHLcHI ∑ −× M M ddHN ϕθϕθυυ )),,(( 0 , (1) where L(H-H 0 ) and N(n-n 0 ) are the form-factors of EPR and ENDOR line shapes, respectively, H 0 and ν 0 are the resonant field of EPR and resonant frequency of EN- DOR, M and m are the projections of electron and nu- clear spins on H direction, C is a constant. In (1) H is a parameter determining the point where EPR spectrum is saturated and, thus, the group of paramagnetic cen- ters with the same definite orientation. H 0 and ν 0 can be obtained from the spin-Hamilto- nian accounting for electron Zeeman, nuclear Zeeman and superhyperfine interactions [21, 25]. ),( ),( ),(0 ϕθβ ϕθυ ϕθ g mAh H mwm − = , (2) =),,(0 ϕθυ HM 2/12 )( ),(                 −        = ∑ ∑ − i ni j ijjj HhAhg hg M υ ϕθ , (3) where 2/1 2)(),(         = ∑ i ii hgg ϕθ , (4) ),(/),( 2/12 ϕθϕθ ghgAA i j jjij                 = ∑ ∑ , (5) hHgH nnn /)( βυ = , (6) and A ij are the components of superhyperfine A-tensor in g-tensor axes frame, h i (i = 1-3) denote the direction cosines of magnetic field vector, ig are the principal values of g tensor, g n is the nuclear g factor, β and β n are the Bohr magneton and nuclear magneton, respec- tively, h is the Planck�s constant, ν mω is the microwave frequency of EPR spectrometer. The expressions for resonant field (2) and resonant frequencies (3) contain many parameters and some sim- plifications can be made. Superhyperfine tensor has been considered to have axial symmetry along the direction on nucleus, r n . Then, in /// ,, zyx axes frame with ||/z r n it can be presented as following:           + − − = ba ba ba A 200 00 00 ' , (7) where ' 3 1 SpAa = , (8) )( 3 1 '2'233 '' AAb −= , (9) and a is a Fermi constant of isotropic superhyperfine interaction, b is a constant of dipole-dipole interaction determining the anisotropic superhyperfine interaction [25]. We shall restrict ourselves to the axial symmetry of g -tensor in EPR. In this case we can consider that ϕ n = 0 and yy ||/ . Then, the superhyperfine tensor in g -tensor axes frame can be written as:           −+ − −+ = )1cos3(0sincos3 00 sincos30)1sin3( 2 2 nnn nnn bab ba bba A θθθ θθθ .(10) Components A ij of A-tensor and ones A i�j� of A�-ten- sor are related by transformation of the axes frame. Superhyperfine interaction between CO 2 - and surround- ing nuclei is smaller than the EPR linewidth, therefore it can be neglected in (2). Besides, the anisotropy of g - tensor of CO 2 - radical in tooth enamel is small [10] and g can be considered as an isotropic value in (3). There- fore, the spectrum of powder ENDOR can be written as: ∫ ∫ −= π π θθυ 0 0 0 2/ ))((sin),( HHLcHI [ ] ϕθϕθυυϕθυυ ddHNHN )),,(()),,(( 00 −+ −+− , (11) [ ] 2/1222 || 2 0 cos)( )( θβ υ θ ω ⊥⊥ −+ = ggg h H m , (12) =± ),,(0 ϕθυ H S. Ishchenko et al.: ENDOR study of irradiated tooth enamel 87SQO, 2(1), 1999 [ ] 2/1 2222 2sin 16 9 )1cos3( 2 1 )(       +−+±= γγυ bba h Hn ,(13) nn θθϕθθγ coscoscossinsincos += , (14) where γ is the angle between magnetic field and r n direc- tion. In our case b << ν n , therefore, the expression (13) can be represented as a series expansion retaining only the first term: ±=± h Hg H nn β ϕθυ ),,(0           −++−+ 1) 2 sin)cos( 2 cos)(cos(3 2 1 222 ϕθθϕθθ nnba h .(15) ENDOR signal defined by (11) is the spectrum aver- aged over the orientations of magnetic field at θ = const (see Fig. 1). If ENDOR frequencies are determined by (15) then the spectrum (11) consists of three doublets cen- tered at ν n . Spectral positions of the doublets peaks are determined by: [ ]{ }1)(cos3 2 1 2 1 −−+±=± θθυυ nn ba h , )0(min == ϕγγ , (16) [ ]{ }1)(cos3 2 1 2 2 −++±=± θθυυ nn ba h , )(max πϕγγ == , (17) 2 ),( 2 1 3 πγυυ =−±=± ba hn . (18) Note, that ν 3 ± is observed only for nθπθ −> 2 . These expressions have been deduced from (11) assuming that the lineshapes are δ-function-like for both EPR and ENDOR lines. Expressions (16)-(18) do not contain the relative intensities of ENDOR doublets and can be used to obtain superhyperfine parameters in the first-order approximation only. Their accurate values can be deter- mined from computer fitting of the experimental spec- trum and the one calculated on the base of (11) or (1). Such analysis has shown that the intensities of υ 3 ± dou- blet exceed considerably the ones of other lines. Thus, υ 3 ± doublet determines ENDOR spectrum if the later takes place. This complicates the study of powder ENDOR because the υ 3 ± doublet has no angular depen- dence. It is convenient to introduce the value +− −= υυδ which does not depend on the shift of υ n at H scan and describes the orientation dependence of ENDOR spec- trum. 4. Experimental results 4.1 EPR EPR spectra of tooth enamel have been studied in de- tail, so we discuss only the data that will be necessary to understand ENDOR results.EPR spectrum observed by us in the unheated samples is determined mainly by the signal from CO 2 - radicals. This signal has been well de- scribed by axial g -tensor with g ⊥ = 2.0021, ||g = 1.9975 and cg |||| ( c is the hexagonal axis of hydroxyapatite) that agrees with our earlier publication [10]. There was no reason to use a rhombic g tensor. The sample annealing resulted in the decrease of the overall intensity of EPR spectrum. It can be explained by complete destruction of disordered CO 2 - radicals [17] and particular disintegration of bulk CO 2 - radicals. Be- sides, the relative intensities of weak signals changed slightly and the new spectrum components revealed. As to new signals, it should be noticed that EPR spectrum consists of seven components centered at g = 2.0003 and separated by 2.18 mT. The central line of this signal is masked by more intensive signal from CO 2 - radical. This spectrum has been assigned to (CH3)-C-R radical acti- vated in organic matter by annealing [11]. EPR spectrum Fig. 1. Definition of the different polar angles describing the magnetic field vector and the direction of the interacting nu- cleus in the g -tensor axes frame. z�� axial axis of the superhy- perfine tensor. S. Ishchenko et al.: ENDOR study of irradiated tooth enamel 8 8 SQO, 2(1), 1999 at g = 2.006 and ∆H pp = 0.1 mT was observed too. It has been attributed to SO 2 - radical [26]. ENDOR sig- nals caused by (CH 3 )-C-R and SO 2 - radicals were not observed. EPR spectra of irradiated tooth enamel be- fore and after the annealing are shown in Fig. 2. The main EPR signal is seen to change slightly under the annealing. 4.2. ENDOR ENDOR spectrum of unheated powder of tooth enamel has been recorded at T = 77 K and T = 4.2 K. It consists of the singlet at Larmor frequency of 31P nuclei and dou- blet at Larmor frequency of 1H nuclei. The intensity of phosphorous line changes according to EPR lineshape while its linewidth does not vary at different magnetic fields within the experimental errors. Notice that field dependence of an ENDOR intensity has been called ENDOR-induced EPR (EI-EPR). The changing of 1H doublet intensity distinguishes slightly from EPR line- shape on the lowfield wing of EPR line. Besides, 1H dou- blet splitting increases considerably on the EPR line wings (Fig. 3). Similar behavior of the 1H doublet field dependence has been also observed by Galtsev [19]. Re- ducing the measurement temperature from 77 to 4.2 K resulted in the increase of the overall intensity of EN- DOR spectra only. The doublet structure of the proton ENDOR signal vanished after annealing. However, a new structure was revealed for both 1H and 31P signals at liquid helium tem- perature and below. This structure was caused by inter- action of CO 2 - radical with nuclei of hydroxyapatite lattice. It is observed for all settings of magnetic field within EPR spectrum and centered at Larmor frequen- cies of 1H and 31P nuclei. The structure of 31P ENDOR spectra consists of nine doublets (see Fig. 4). The δ values for all doublets are represented in Table I. The peaks of each ENDOR dou- 336 340 b a E P R s ig n a l, a rb .u n . Magnetic field, mT Fig. 2. EPR spectrum of irradiated tooth enamel powders near g = 2.0, T = 300 K: a � unheated sample, b � sample dried at T = 2500C during 45 minutes. 5 6 12 13 14 15 16 1 H 31 P H= 329.8 mT H=330.7 mT H=331.3 mT E N D O R s ig n a l, a rb .u n . Radiofrequency, MHz Fig. 3. ENDOR spectra of unheated sample of tooth enamel for different settings of magnetic field, υ mω= 9252 MHz. 5,5 6,0 ν, MHz θ = 53 0 H=332.1 mT 9 8 7 6 5 4 3 2 1 x4E N D O R s ig n a l, a rb .u n . Fig. 4. 31P ENDOR spectrum of annealed powder of tooth enamel. ENDOR doublets caused by nuclei of different shells are labeled by 1-9. S. Ishchenko et al.: ENDOR study of irradiated tooth enamel 89SQO, 2(1), 1999 blet can be formed by nuclei of one shell as well as by the contributions of nuclei involved into several shells. The structure of 31P ENDOR spectra can be arbitrarily sub- divided on a three groups of lines. The first group contains the doublets 1, 3, 6, 7 which depends on setting of the magnetic field. The second group involves dou- blets 2, 4, 5. Their spectral positions do not shift when a setting of the magnetic field is changed. Finally, the lines being near Larmor frequency are attributed to the third group. These are 8 and 9 doublets; their lineshapes and intensities are modified at different settings of the mag- netic field. But their spectral shifts are small and can not be determined accurately. The structure of 1H ENDOR spectra is less distinc- tive than the one 31P although seven doublets can be ex- uded. The typical 1H ENDOR spectrum is represented in Fig. 5. Its lineshape is modified while the setting of the magnetic field is changed, however, the shifts of the peaks can not be determined accurately. The values of δ are represented in Table II. Note that intensive struc- tureless signal centered at υ n revealed in both 1H and 31P ENDOR spectra too. It can be attributed to centers located in sites with high concentration of defects. The contribution from distant nuclei is also possible. 5. Analysis of the spectra and discussion The admixed hydrogen atoms other than ones involved into hydroxyapatite lattice are present in unheated tooth enamel. They occur in crystallites in the forms of H 2 O molecules, admixed OH groups and H-centers of differ- ent types. These atoms form the large amounts of hy- droxyapatite defects that lead to disappearance of supe- rhyperfine structure in ENDOR spectra. Besides, the hy- drogen arranged randomly appears to be responsible for Table I. 31 P superhyperfine coupling constants of the CO 2 - radical in tooth enamel. Doublet 1 2 3 4 5 6 7 8 9 δ, kHz 920- 860 530 380-320 220 170 130-110 60-40 40-30 11 h/a , kHz -310 ±20 -165 a ±20 -90±20 -49 a ±15 -33 a ±10 -20±10 0±10 0 h/b , kHz 630 ±20 385 ±20 260 ±20 183 ±15 147 ±10 100 ±10 60 ±10 56- 38b θn, deg. 28±4 90±4 40±4 90±4 90 ±4 60 ±4 65 ±4 rn, nm 0.37 ±0.01 0.44 ±0.03 0.50 ±0.01 0.56 ±0.03 0.60 ±0.03 0.68 ±0.01 0.81 ±0.01 l nθ , deg 33 90 47 90 90 59 65 l nr , nm 0.41 0.47 0.50 0.55 0.63 0.68 0.80 0.83- 0.94 0.97 � shell I II III IV V VI VIII IX-XIV XV,. a � values calculated with using the expression (19) b � values calculated with using the expression (20) and l nr Fig. 5. 1H ENDOR spectrum of annealed powder of tooth enamel. ENDOR doublets caused by nuclei of different shells are numbered by 1-7. 13,5 14,0 14,5 θ = 800 H=331.9 mT 7 6 5 4 3 2 1 ν, MHz x3 EN D O R si gn al , a rb .u n. S. Ishchenko et al.: ENDOR study of irradiated tooth enamel 9 0 SQO, 2(1), 1999 the doublet structure of proton ENDOR signal. The sim- ilar spectrum lineshape has been described in the work [24] devoted to matrix ENDOR. The peaks positions of such spectrum are determined by competition between signal increase from distant nuclei due to the growth of their amounts and signal decrease due to the diminution of the transition probabilities and the relaxation mech- anisms. The change of the doublet splitting of the proton ENDOR (Fig. 3) and deviation of EI-EPR lineshape from EPR one can be explained by the presence in the ENDOR signal of two components corresponding to two paramagnetic centers. The first component is a narrow intensive doublet with the splitting ∆ 1 = 100-140 kHz. It dominates in the central part of the EPR spectrum. The value of ∆ 1 is changed within the mentioned region for different samples but it is constant for a certain sample between g = 1.9975 and g = 2.0021. Thus, the narrow component can be attributed to the main EPR signal caused by the bulk CO 2 - radical. The second component is the wider doublet with splitting ∆ 2 = 260 kHz. It dom- inates on the wings of the EPR spectrum. The nature of a center that causes this signal is not clear. The above- mentioned disordered CO 2 - radical is probably respon- sible for it. The relaxation characteristics of the disor- dered radicals appear to be more favourable for EN- DOR observations. This increases the contribution of these centers in EI-EPR signal and results in different lineshapes of EPR and EI-EPR spectra. Such relaxation amplification of ENDOR signal is probably absent in phosphorous ENDOR. ENDOR spectra of the annealed tooth enamel samples have been analyzed using the the- ory of powder ENDOR. The a , b and θ n parameters have been estimated using (16) - (18) and the assump- tion that ( ) ( )[ ] 2/122 || 22 ⊥⊥ −−= ggggθ where Hhg m βυ ω /= at H = H 0 . Accurate values of parame- ters have been determined by computer fitting of the spectra calculated according to (11) to experimental ones. In these simulations the gaussian with ∆H = 0.25 mT and lorentzian with ∆υ = 10-30 kHz for different shells have been used as EPR and ENDOR lineshapes, respec- tively. Note that parameters variation did not resulted in the accurate agreement between calculated and experi- mental dependencies. Experimental curves were more smooth than the calculated ones. The rhombic g-tensor with g x = 2.0030, g y = 2.0015, g z = 1.9970 represented in [16] and the expression (1) have been used to simulate the ENDOR spectra, too, but improvement has not been obtained. Thus, the mentioned discrepancy appears to be due to the neglection of relaxation orientation de- pendencies in (1) and (11). The analysis of the expressions (16)-(18) and (11) has shown that angular variation of the second group (dou- blets 2, 4, 5) of phosphorous ENDOR is characteristic for nuclei with θ n = 900. Powder ENDOR spectra of such nuclei consist of dominating doublets (18) which have no orientation dependencies. In this case the value ba − can be determined at once from the experiment. To esti- mate the constants a and b separately the following ap- proximation can be made: kreaa −= 0 , (19) where 0a and k are the parameters which can be deter- mined using the expression (19) for nuclei of first group (doublets 1, 3, 6, 7). The anisotropic superhyperfine interaction has been considered as pure dipole-dipole interaction. Then constant b can be written as: 3r gg b nn ββ = , (20) where r ≡ r n corresponds to the distance between CO 2 - radical site and nucleus n. The values θ n and r n obtained from ENDOR spectra for different nuclei allow to determine the location of CO 2 - radical in crystal lattice. The values l nθ and nr for differ- ent possible sites of CO 2 - in hydroxyapatite lattice have been calculated using the model of its structure (Ref. 28) and compared with θ n and r n . It has been found that only CO 2 - in B site (phosphorous substitution) at cg |||| agrees well with the experimental data. Some discrepancy between experimental and lattice data for phosphorous nuclei of shell I (see Table I) can be explained by their shift towards the defect. Note that the ENDOR lines due to 31P nuclei of VII shell were not observed. In accor- dance with the computer simulation these lines have in- tense angular dependencies and they are broadened strongly. Besides, they are likely superimposed with lines of shell V and VI. A constant a can be considered to be equal to zero for phosphorous of VIII shell and more distant nuclei. Then the frequency region for lines caused by these nu- clei can be estimated by means of (15) and (20) using lattice data for l nθ and l nr . Such calculations permit to explain the peaks 8, 9 in phosphorous ENDOR (see Table I). The values l nθ and l nr have been also used to simulate proton ENDOR spectra. Constants b have been calculated from l nr . Constants a have been determined by computer fitting of calculated ENDOR spectra to experimental ones. The obtained values of a have not been contradicted with the model of center when OH vacancy is situated in the nearest surrounding of CO 2 - radical. This finding agrees with the well-known model [29] of carbon incorporation in hydroxyapatite structure: OHCa VCOVOHPOCa ++↔++ −−−+ 3 3 3 4 2 , (21) where CO 3 3- molecular ion is a precursor of CO 2 - radical. The CO 2 - site and surrounding apatite structure are represented in Fig. 6-8. The neighbour nuclei 31P and 1H are numbered in accordance with Tables I, II. S. Ishchenko et al.: ENDOR study of irradiated tooth enamel 91SQO, 2(1), 1999 6. Conclusion The signal caused by CO 2 - dominates in EPR spectrum of irradiated tooth enamel. The contributions of other centers have been revealed too, but their intensities are smaller. Two types of CO 2 - radicals are observed in un- heated samples. One of them is the bulk center. It causes the angular variations of EPR spectrum in a tooth plate. The centers of the second type are disordered and cause EPR spectrum of powder line shape. They appear to be located on the surface of crystallites. The singlet at Larmor frequency of 31P nuclei and doublet at Larmor frequency of 1H nuclei are observed in ENDOR spectrum of unheated tooth enamel. The value of doublet splitting depends on the settings of mag- netic field and is equal to 100-140 kHz and 260 kHz in the central part and wings of EPR line, respectively. The doublet structure of proton ENDOR can be explained by the theory of matrix ENDOR assuming the presence of a large amount of admixed hydrogen atoms. The modi- fication of proton ENDOR line at different settings of magnetic field appears to be connected with the exist- ence of two types of CO 2 - radicals. These radicals cause the proton ENDOR signal with various splittings, but their contributions are different in the central part and on the wings of EPR line. The annealing at 250 0C results in the decrease of the overall intensity of EPR spectrum due to complete de- struction of the disordered CO 2 - radicals and particular disintegration of the bulk CO 2 - radicals. Small variation of EPR lineshape due to the redistribution of the inten- sities of weak signals and appearance of the new signal from centers activated by annealing is also observed. The structure of phosphorous and proton ENDOR centered at Larmor frequencies of 1H and 31P nuclei reveals after annealing. Its analysis allows to determine the constants of superhyperfine interaction of CO 2 - radical with neigh- bour nuclei and to conclude that CO 2 - ion substitutes phosphorous in hydroxyapatite lattice of tooth enamel (B site). This substitution is found to be accompanied by the creation of OH vacancy in the nearest surroundings and shift of phosphorous nuclei of shell I towards the Fig. 6. Fragment of hydroxyapatite structure. δγβα ,,, are the hexagonal axes. Fig. 7. Hydroxyapatite structure close to the hexagonal c-axis and proposed site for CO 2 - radical. The numbers I-X are shown the location of protons of different shells. Fig. 8. The locations of phosphorous nuclei of different shells relative to CO 2 - radical. S. Ishchenko et al.: ENDOR study of irradiated tooth enamel 9 2 SQO, 2(1), 1999 defects by dr = 0.04 nm. The experimental angular de- pendencies of ENDOR lines are more smooth than the calculated ones. This discrepancy appears to be due to the neglection of an orientational dependence of relaxation times in the theory. All EPR and ENDOR spectra have been described by axial g -tensor with ||g = = 1.9975 and ⊥g = 2.0021 at cg |||| . Acknowledgments The authors wish to thanks Prof. A. B. Roitsin for con- sultations on the theory of powder ENDOR and Dr. S. V. Virko for assistance in computer calculations. References 1. M. A.Scott, H. G.Hallacher, T. P. J. Han, B. Henderson, Radiat.Eff.Defect S., 136, p.957 (1995). 2. E. Borovsky, Terapevticheskaya Stomatologiya, Meditsina, Mos- cow, 1988 (in Russian). 3. J. Tatsumi-Miyajima, Nucl.Instr.&Meth., A257, p.417 (1987). 4. M. Ikeya, Appl.Radiat.Isot., 44, p.1(1993). 5. A. B. Brik, S. S. Ishchenko, L. G. Rosenfeld, D. I. Zabolotnii, I. S. Zaritskaya, Med.Radiology, p.25(1993). 6. R. Grun, Radiat.Isot., 40, p.1045 (1989). 7. B. A. Blackwell, H. P. Schwarcz, Appl.Radiat.Isot., 44,p. 243 (1993). 8. K. Sato, Calcif.Tissue Int., 29, p.95 (1979). Doublet 1 2 3 4 5 6 7 δ, kHz 1350 650 370 280 185-175 140 62 θn l , deg. 60 78,37 70,33 51 46,81,75 rn l , nm 0.42 0.60 0.62,0.67 0.75 0.81,0.81,0.83 0.9,� kHz,/1~/ 3 nrhb 1070 367 332 188 155,149,138 kHzh,/a -280 ±30 -20 ±20 0 ±20 0 ±20 0 ±20 0 shell II (III-IV) III-V(VI) V,VI,(VIII- IX) VII VIII,IX,X XI,� Table II. 1 H superhyperfine coupling constants of the CO 2 - radical in tooth enamel. 9. Y. Doi, T. Aoba, M. Okazaki, J. Takahashi, Y. Moriwaki, Calcif. Tissue Int., 28, p.107 (1979). 10. S. S. Ishchenko, S. M. Okulov, I. P. Vorona, A. B. Roitsin, A. A. Klimov, Sov. Phys. Solid State, 34, p.1522 (1992). 11. G. Baquet, V. Q. Truong, M. Vignoles, J. C. Trombe, G. Bonel, Calcif. Tissue Int., 33, p.105 (1981). 12. F. J. Callens, R. M. H. Verbeeck, P. F. A. Matthys, L. C. Mar- tens, E. R. Boesman, Calcif. Tissue Int., 41, p.124 (1987). 13. F. J. Callens, Nucleonika, 42, p.565 (1997). 14. P. Moens, F. Callens, S.V an Doorslaer, P. Matthys, Phys. Rev., B53, p.5190 (1996). 15. T. Shimano, M. Iwasaki, C. Miyazawa, T. Miki, A. Kai, M. Ikeya, Appl.Radiat.Isot., 40, p.1035 (1989). 16. P. D. Moens, F. J. Callens, P. F. Matthys, J. Chem. Soc. Faraday Trans., 90, p.2653 (1994). 17. F. Callens, P. Moens, P. Verbeeck, Calcif.Tissue Int., 56, p.543 (1995). 18. H. van Willigan, A. H. Roufosse, M. J. Glimcher, Calcif.Tissue Int., 31, p.70 (1980). 19. V. E. Galtsev, Appl. Radiat. Isot., 47, p.1365 (1996). 20. G. Rist, J. Hyde, J. Chem.Phys., 52, p.4633 (1970). 21. G. C. Hurst, T. A. Henderson, R. W. Kreilick, J.Am.Chem.Soc., 107, p.7294 (1985). 22. P. A. Narayana, M. K. Bowman, D. Becker, L. Kevan, J.Chem.Phys., 67, p.1990 (1977). 23. B. M. Hoffman, R. A. Venters, J. Martinsen, J.Magn.Res., 62, p.537 (1985). 24. L. Kevan, S. Schlick, K. Torlyama, M. Iwasaki, J.Phys.Chem, 84, p.1950 (1980). 25. V. G. Grachev, M. F. Deigen,Usp. fiz. nauk, 125, p.631 (1978). 26. D. U. Schramm, A. M. Rossi, Appl.Radiat.Isot, 47, p.1443 (1996). 27. J. R. Niklas, J.-M. Spaeth, Phys.Stat.Sol.(b), 101, p.221 (1980). 28. M. I. Kay, R. A. Young, A. S. Posner, Nature, 204, p.1050 (1964). 29. J. C. Larbarthe, M. Therasse, G. Bonel, G. Montel, hebd. Seanc.Acad.Sci.Paris,C, 276, p.1175 (1973).