- Email: [email protected]
EPR dosimetric properties of formates
ARTICLE IN PRESS
Applied Radiation and Isotopes 59 (2003) 181–188
EPR dosimetric properties of formates Tor Arne Vestada, Eirik Malinena, Anders Lundb, Eli Olaug Holea, Einar Sagstuena,* a
Department of Physics, University of Oslo, P.O. Box 1048 Blindern, N-0316 Oslo, Norway b Chemical Physics Laboratory, IFM, Linkoping University, S-581 83 Linkoping, Sweden . . Received 2 April 2003; received in revised form 7 May 2003; accepted 22 June 2003
Abstract As a part of a program to develop an electron paramagnetic resonance (EPR) dosimeter suited for clinical use (doses in the cGy range), polycrystalline samples of lithium formate monohydrate (HCO2Li H2O), magnesium formate dihydrate (C2H2O4Mg 2H2O), and calcium formate (C2H2O4Ca) have been examined. l-Alanine was included for comparison and reference. Samples were irradiated with 60Co g-rays and 60–220 kV X-rays. The dosimeter response was assessed using the peak-to-peak amplitude of the first-derivative EPR spectrum. Dose–response curves for the 60Co g-irradiated samples were constructed, and the dependences of the response on the photon energy, microwave power, and modulation amplitude were studied. Stability of the irradiation products upon storage (signal fading) was also investigated. Lithium formate monohydrate is by far the best candidate of the tested formates, suitable for measuring doses down to approximately 0.1 Gy. Lithium formate monohydrate is more sensitive than alanine by a factor of 5.6– 6.8 in the tested photon energy range, it exhibits no zero-dose signal and shows a linear dose response in the dose range from 0.2 to 1000 Gy. Its EPR signal was found unchanged in shape and intensity 1 week after irradiation to 10 Gy. Various less favorable properties rendered the other formates generally unsuitable, although calcium formate exhibits some interesting EPR dosimetric properties. r 2003 Elsevier Ltd. All rights reserved. Keywords: EPR; EPR dosimetry; Lithium formate; Formates; Dosimetry; Therapeutic range
1. Introduction Electron paramagnetic resonance (EPR) dosimetry with l-alanine has become a standard technique at the IAEA for measurements of high doses and for transfer dosimetry (Nette et al., 1993; Mehta and Girzikowsky, 1995, 2000). However, even with several decades of EPR studies on alanine (Bradshaw et al., 1962; Regulla and Deffner, 1982), the basic properties of alanine are still under investigation (Callens et al., 1996; Sagstuen et al., 1997; Vanhaelewyn et al., 1999; Heydari et al., 2002; Malinen et al., 2003). In addition, searches for alternatives to alanine as a dosimetric material are *Corresponding author. Tel.: +47-228-55653; fax: +47-22855671. E-mail address: [email protected] (E. Sagstuen).
actively pursued (Bogushevich and Ugolev, 2000; Hassan and Ikeya, 2000; Ikeya et al., 2000; Olsson et al., 2000; Murali et al., 2001; Lund et al., 2002). An important motivation for such searches is that the sensitivity of alanine dosimetry is too low to make this technique a realistic alternative to the dosimetry systems currently in use in clinical practice, where doses down to 0.1 Gy should be easily detectable. Radiation-induced paramagnetic defects in formates (salts of formic acid, HCOOH) have been proposed as dose indicators for application in EPR dosimetry (Lund et al., 2002). In the present study, which is a part of a program to develop an EPR dosimeter for doses in the cGy range, polycrystalline samples of lithium formate monohydrate (HCO2Li H2O), magnesium formate dihydrate (C2H2O4Mg 2H2O), and calcium formate (C2H2O4Ca) have been examined. Pellets of the samples
0969-8043/$ - see front matter r 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0969-8043(03)00165-9
ARTICLE IN PRESS 182
T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188
were irradiated using 60Co g-rays and X-rays with energies between 60 and 220 kV, and the dependences of the EPR spectra of the various formates on microwave power and modulation amplitude were studied. The dose–response curves for the 60Co g-irradiated samples were constructed in the 0.5–1000 Gy dose range (lithium formate monohydrate was additionally examined in the dose range from 0.2 to 1.0 Gy). Furthermore, for each of the studied formates, the dependence of the EPR spectrum intensity on photon energy was investigated using the intensity of the alanine spectrum as a reference, and compared with the corresponding calculated mass energy-absorption coefficients. Finally, the stability of the irradiation products upon storage (signal fading) was studied. Pilot experiments using sodium formate (CHO2Na), potassium formate (CHO2K), ammonium formate (CH5NO2), and cesium formate (CHO2Cs) were also performed. Potassium formate was not examined further, as it was realized that utilizable pellets could not be made because of the hygroscopic properties of this material. In the case of sodium formate, a relatively large hyperfine coupling (hfc) from sodium was observed (B0.8 mT), which is unfavorable for EPR dosimetry, and the sensitivity was low as compared with the other formates tested. Ammonium formate is, however, a promising material for sensitive EPR dosimetry (Gustafsson et al., 2003). Cesium formate gave no detectable EPR signal after irradiation with Xrays (100 kV) to the dose of 100 Gy, apparently in contradiction to the previous results (Supe et al., 1986). The main signal observed in the spectra of irradiated formates is attributed to CO 2 radicals (Ovenall and Whiffen, 1961; Orsega and Corvaja, 1977; Supe et al., 1986; Lund et al., 2002). The EPR spectrum of CO 2 radicals is single-lined, making the polycrystalline spectrum from the formates fairly simple, mainly dependent on the strength of the hfc interaction with the alkali metal cation and the g-tensor anisotropy. The spectrum of HCO2Li H2O pellets is a single line with the width approximately 1.5 mT. In the other investigated formates, the interaction of the unpaired electron with the cation nucleus is weaker, the g-tensor anisotropy is resolved (under non-saturating conditions and low modulation amplitudes), and the lines are, accordingly, narrower (approximately 0.25 mT).
2. Materials and methods 2.1. Samples Polycrystalline lithium formate monohydrate (HCO2Li H2O) was obtained from Sigma-Aldrich; magnesium formate dihydrate (C2H2MgO4 2H2O) and calcium formate (C2H2CaO4) were obtained from Fluka. These
materials are hereafter referred to as Li-Mg- and Caformate. Samples were made in the form of cylindrical pellets (D ¼ 5 mm, h ¼ 471 mm) with a manual pellet press (10-kN force was applied for 1 min). Alanine pellets were obtained from Bruker BioSpin (Bruker Bronze lot No. 587895/1). The polyethylene content of the alanine pellets (20% by weight) was accounted for in the analysis of the signal intensities. During irradiation and storing, the pellets were kept in cylindrical polyethylene vials (thickness 1 mm, inner diameter 6 mm), arranged in groups three by three. The temperature and relative humidity in the laboratory during preparation, storing, and EPR measurements were 2572 C and 1572%, respectively. The pellets were stored in darkness and exposed to luminescent light only when transferred from the storage area to a radiation source or the EPR spectrometer.
2.2. Radiation The X-ray unit used was a Pantak HF225 generator, with a maximum voltage and power of 225 kV and 3.2 kW. The inherent filtration was 1 mm beryllium, and two primary filters were available, namely, 1.5 mm aluminum and 0.5 mm copper. The used combinations of voltages and filters are listed in Table 1, along with the corresponding half value layers (HVLs) in aluminum. HVL is defined as the thickness of an absorber that reduces the air kerma of an X-ray beam by 50% (IAEA, 1987, 2000). During the X-ray irradiation, vials with pellets were sitting on a 10-mm-thick PVC slab about 0.5 m away from the source. The 60Co g-irradiation was performed with a Mobaltron 80 therapy unit (TEM Instruments Limited). The vials with the pellets were located 0.80 m far from the source with a 5-mm buildup of solid water and over 100 mm of solid water as a backscatter material. In both X-ray and 60Co irradiations, electronic equilibria were achieved. All the quoted radiation doses are expressed as doses to alanine, since alanine is chosen as the reference medium in this work. For theoretical estimates of the energy response, the mass energy-absorption coefficients and the effective
Table 1 Radiation beams with the filters used and the corresponding HVL employed in this study Beam
Filter
HVL (mm Al)
60 kV 100 kV 220 kV 60 Co g-rays
1.5 mm Al 1.5 mm Al 1.5 mm Al+0.5 mm Cu —
1.6 3.0 15 42.3
ARTICLE IN PRESS T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188
atomic numbers of the samples were calculated using program XMuDat (Nowotny, 1998). 2.3. EPR measurements EPR spectra were recorded at room temperature using a Bruker ESP300E spectrometer equipped with a standard X-band bridge and a dual rectangular cavity (TE104). Upon measurements, the pellet was placed in an open quartz tube with an inner diameter of 5.25 mm, and a sample support system was used to ensure identical and reproducible positions of all samples in the cavity. A Mn2+/MgO reference sample (JEOL Co.) was placed into the second cavity, and the EPR spectrum of Mn2+ was recorded immediately after each dosimeter spectrum registration in order to correct the spectrum intensity for spectrometer instabilities. In all measurements except those in the fading experiments (see below), the EPR signal from the reference sample was recorded using a microwave power 5 mW and a modulation amplitude 0.15 mT. All reference sample measurements showed a relative standard deviation of less than 1.2%. The term ‘EPR intensity’ corresponds to the peak-to-peak amplitude of the first-derivative curve of the EPR spectrum. In the studies of the signal intensity dependences on microwave power ðPÞ and modulation amplitude (MA), one sample of each material, irradiated with 60Co g-rays to 10 Gy, was used. The microwave power was being decreased from 202 to 1 mW (0–23 dB) in steps of 1 dB, and the EPR spectrum was recorded at each step. This stepwise power change was repeated for 20 modulation amplitude settings increasing from 0.1 to 3.1 mT. Thus, 480 measurements in total have characterized the P=MA dependence profile for each material (only selected data are presented in the figures of this paper). Based on the P=MA dependence profiles, individual microwave power and modulation amplitude settings were chosen
Table 2 The individual microwave power ðPÞ and modulation amplitude (MA) settings employed in this study with the resulted EPR intensities (expressed in percent of the maximum intensity achievable in the ranges 1–202 mW and 0.1–3.1 mT) and the time between the sample irradiation and the spectrum registration
Alanine Li-formate Mg-formate Ca-formate
183
Table 3 Average relative standard deviations of the signals from replicate pellets of various dosimetric materials irradiated to the same dose Dose range (Gy)
Alanine
Li-format
Mgformate
Caformate
1.0–10 0.2–1.0
1.9% —
1.2% 4.3%
2.7% —
2.7% —
for each material (Table 2) that produce signals equal to 5271% of its maximal value (marked in Fig. 2) in the P/MA ranges investigated. This approach makes it possible to compare performance characteristics of different dosimetric materials accurately. The measurements related to the spectrum shape (60Co, 100 Gy), energy response (10 Gy), and dose response (60Co) were performed 64–70 h after irradiation. Each data point in Figs. 3 and 5 represents the average of the signals of three similar samples irradiated simultaneously; the error bars (barely visible) represent one standard deviation of the mean. Additional dose response measurements for subgray doses were performed for Li-formate, with five samples irradiated simultaneously for each dose point (Fig. 4). Average relative standard deviations for the various dosimeters are listed in Table 3. In the study of fading, one sample of each material irradiated to 10 Gy with 220 kV X-rays was used. The temperature during irradiation was 295 K. Its signal measurements were made at equal time intervals (ranged between 9 and 40 min for different materials) continuously for 6–7 days, starting 30–45 min after the end of the irradiation. The room temperature and humidity did not vary significantly during the measurement sessions. In the Mg-formate and alanine fading experiments, the reference sample signals were recorded at the 10 mW power and the 0.47 mT modulation amplitude, in contrast to the parameters specified above. No systematic trends in the reference sample amplitudes were detected that would indicate, e.g., a significant accumulation of moisture in the sample pellets during the timeextended measurement series.
3. Results
Power (mW)
MA (mT)
EPR intensity (%)
Time (h)
3.1. EPR spectra of
6.33 39.9 20.0 15.9
0.990 1.25 0.990 0.990
51.8 51.2 52.0 52.8
64 66 68 70
EPR spectra of the pellets of alanine, Li-, Mg- and Ca-formates irradiated to 100 Gy (60Co g-rays) are shown in Fig. 1. The alanine spectrum consists of the contributions from at least three different radical species (Sagstuen et al., 1997), while the relatively sharp
60
Co g-irradiated samples
ARTICLE IN PRESS T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188
184
3.2. Dependence of the EPR spectra on microwave power and modulation amplitude Alanine 1.0
Li-formate 5.6
The power and modulation amplitude dependence profiles for alanine, Li-, Mg-, and Ca-formate are shown in Fig. 2. The pellets were irradiated to 10 Gy using 60Co g-rays. The alanine profile shows a clear maximum at 50 mW and 1.6 mT. The profiles of the formates do not feature any well-defined maxima. The highest intensities, however, were achieved at the parameter values listed in Table 4. 3.3. Dose–response curves for
Mg-formate 4.2
Ca-formate 4.3 342
344
346
348
350
352
B-field (mT) Fig. 1. First-derivative EPR spectra of pellets of alanine, Li-, Mg-, and Ca-formate irradiated to 100 Gy with 60Co g-rays. The upper spectrum for each compound was recorded at P ¼ 4:0 mW and MA=0.1 mT, the lower one was recorded at the individual power and modulation amplitude settings listed in Table 2. All the spectra are normalized to the intensity of the alanine spectrum, and the normalization factors for the lower spectra are quoted under each curve. The g value at the center of each spectrum is 2.0023.
resonance from the irradiated formates has been attributed to CO 2 radicals (Ovenall and Whiffen, 1961; Orsega and Corvaja, 1977; Supe et al., 1986; Lund et al., 2002). The EPR spectra of Li-, Mg-, and Caformate are different in shape due to the different nuclear spins, nuclear g-factors and isotopic abundances of the cations (7Li 92.5%, I ¼ 3=2; gN ¼ 5:792; 6Li 7.5%, I ¼ 1; gN ¼ 0:822; 25Mg 10.13%, I ¼ 5=2; gN ¼ 0:342; 43Ca 0.14%, I ¼ 7=2; gN ¼ 0:376), and due to different anisotropies. Li-formate has an apparently symmetrical 1.5-mT wide resonance line ðg ¼ 2:0005Þ: The spectrum shape is not sensitive to the power and modulation amplitude settings (Fig. 1). The spectra of Mg- and Ca-formate are very similar to each other, and they exhibit the characteristic doublet-like pattern of the powder spectrum of CO 2 radicals well known from the literature (Callens et al., 1995, 1998). High power and modulation amplitude settings mask the doublet.
60
Co g-irradiated samples
Fig. 3 shows the dose–response curves for 60Co girradiated samples in the ranges 0.5–1000 Gy (Fig. 3a, logarithmic scale) and 0.5–10 Gy (Fig. 3b, linear scale). The individual EPR settings producing 52% of the maximal EPR intensities were used (Table 2). Li- and Ca-formate show a linear dose response in the dose range from 0.5 to 1000 Gy without any zero-dose signal. Mg-formate exhibits a substantial zero-dose signal (equivalent approximately to 1 Gy), which is different from the radiation-induced signal, and the dose dependence of the measured EPR signal intensity is not linear below 2 Gy. Alanine generally exhibits a zero-dose signal as well, and its shape and intensity depend somewhat on the production procedure (Schaeken and Scalliet, 1996). Results for Li-formate in the dose range 0.2–1.0 Gy, including the EPR spectra, are presented in Fig. 4. Liformate shows a linear dose response in this dose range. Reinsertion of the same pellet irradiated to 0.4 Gy into the spectrometer cavity 10 times resulted in a statistical sample of the signal values with the relative standard deviation of 4.5% (the data are not shown). 3.4. Effect of the photon energy Fig. 5 shows the dependence of the dose responses of the studied materials on the photon energy. The intensities of the signals of each material were normalized to the intensity of the signal of alanine irradiated to the same dose with the photons of the same energy. The qualities of the used photon beams expressed as HVL in Al are listed in Table 1. The individual EPR settings producing 52% of the maximal EPR intensities were used (Table 2). It appears from Fig. 5 that Li-formate is more sensitive to the photon energy than alanine by a factor of 5.6–6.8 in the tested radiation quality range (HVL from 1.6 to 42.3 mm in Al). Furthermore, of the materials tested, Li-formate appears to be the least sensitive in its response (relative to alanine) to the photon energy. Mg-formate exceeds alanine in the sensitivity to the photon energy by a factor of 4.2–8.2.
ARTICLE IN PRESS T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188 MA (mT)
100 1.57 1.97 1.25 2.49 0.99 0.79 3.13 0.62
75 50 25
0.39
MA (mT)
Li-formate EPR intensity (rel. units)
EPR intensity (rel. units)
Alanine
185
100
3.13 2.49 1.97
75
1.57 1.25
50
0.99 0.79 0.62
25
0.39 0.20
0.20
0 0
100 150 Power (mW)
200
0
100
150
200
Power (mW) MA (mT)
Mg-formate EPR intensity (rel. units)
05
2.49 1.97 3.13 1.57 1.25
100 75
0.99 0.79
50
0.62 0.39
25
MA (mT)
Ca-formate EPR intensity (rel. units)
05
1.97 2.49 1.57 3.13 1.25
100
0.99
75
0.79 0.62
50
0.39
25 0.20
0.20
0 05
0
100
150
200
Power (mW)
05
0
100
150
200
Power (mW)
Fig. 2. Dependence of the peak-to-peak intensities of the EPR spectra of alanine, Li-, Mg-, and Ca-formate irradiated to 10 Gy (60Co) on microwave power (1–202 mW) and modulation amplitude (0.1–3.1 mT, selected data). The individual microwave power and modulation amplitude settings chosen for the experiments in this work are marked in the figure. Each profile is scaled to make the maximal intensity equal to 100.
Table 4 The EPR spectrometer settings corresponding to maximum EPR intensity achievable in the ranges 1–202 mW and 0.1–3.1 mT
Alanine Li-formate Mg-formate Ca-formate
Power (mW)
MA (mT)
EPR intensity (%)
50 202 202 202
1.6 3.1 2.0–3.1 2.0–3.1
100 100 98–100 98–100
Ca-formate is about 27 times more sensitive than alanine at low photon energies, but the relative sensitivity decreases to the 4.3-fold for 60Co g-radiation. 3.5. Mass energy-absorption coefficients The effective atomic numbers of alanine, Li-, Mg-, and Ca-formate are 6.78, 7.31, 8.64, and 14.5, respectively, while the corresponding characteristic for water is 7.51 (Nowotny, 1998). The calculated dependences of the mass energy-absorption coefficients relative to water on photon energy in the range from 10 keV to 10 MeV
are shown in Fig. 6. The dependence is the weakest for Li-formate (the coefficients change between 0.84 and 0.93), while the variation ranges for alanine, Mg- and Ca-formate are 0.68–0.97, 1.50–0.93, and 7.85–0.91, respectively. 3.6. Stability of the irradiation products (signal fading characteristics) Fig. 7 shows the changes in the intensities of the EPR spectra of alanine, Li-, Mg-, and Ca-formate in time after irradiation to 10 Gy with 220 kV X-rays. The intensities at the start of the measurements (30–45 min after end of irradiation) are set to 100%. The signals have been corrected for variations in the Mn2+ reference signal. The large scatter in the results for alanine as compared with those for the formates is due to the relatively weak signals obtainable from alanine at this dose. The EPR signal of Li-formate remains unchanged in shape and intensity 1 week after irradiation to 10 Gy. The signal of Mg-formate fades to approximately 80% of the initial intensity in 7 days. The signal of Caformate increases by approximately 3% in the first few hours, but subsequently fades to approximately 75% of the initial intensity in 7 days.
ARTICLE IN PRESS T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188 1000
100
100 10
EPR intensity (arb. units)
Alanine Li-formate Mg-formate Ca-formate Regression
1 0.1 1
(a) EPR intensity (arb. units)
Li-formate Regression
80 60 40 20 0
0.1
10 100 Dose (Gy)
1000
Alanine Li-formate Mg-formate Ca-formate Regression
8 6
0 2
4 6 Dose (Gy)
0.4 0.6 Dose (Gy)
8
10
Fig. 3. Dose dependence of the peak-to-peak intensities of the EPR spectra of alanine, Li-, Mg-, and Ca-formate irradiated with 60Co g-rays. Individual microwave power and modulation amplitude settings were used for each compound (Table 2).
b)
0.8
1.0
0 Gy 0.2 Gy 0.4 Gy 0.6 Gy 0.8 Gy 1.0 Gy
Li-formate
2
0
0.2
40
4
(b)
0.0 (a)
EPR intensity (arb. units)
EPR intensity (arb. units)
186
20
0
-20
-40
347
348
349
350
B-field (mT)
4. Discussion 4.1. General The different materials tested in this study all exhibit linear dose responses in the dose range tested, except of Mg-formate, whose significant zero-dose signal results in a deviation from linearity at the lowest doses. The spectrum of Li-formate has the most convenient shape from the viewpoint of EPR dosimetry, as the spectrum appears less sensitive to the microwave power and modulation amplitudes in the tested ranges. The effect of the atomic number on the absorption of photons with energies below 250 keV is demonstrated experimentally and theoretically in Figs. 4 and 5. The response of alanine itself is clearly dependent on the photon energy (Fig. 6) due to a relatively low effective atomic number as compared with water, which makes the use of this material in soft X-ray dosimetry less straightforward. In the present work, individual power and modulation amplitude (P/MA) settings were used for each material, they are listed in Table 2. These settings provide equal EPR intensities for each material relative to the maximal achievable values, which were determined from the
Fig. 4. (a) Dependence of the peak-to-peak intensity of the EPR spectrum of Li-formate on the absorbed dose to alanine in the sub-gray range. 60Co g-rays. Five replicate samples for each dose. (b) Examples of the corresponding first-derivative EPR spectra. Microwave power 39.9 mW, modulation amplitude 1.25 mT, 10 scans, total acquisition time 7 min. The background signal from the cavity has been subtracted from the spectra.
profiles shown in Fig. 2. This appears to be a necessary approach when different materials need to be compared with each other in terms of their performance characteristics in EPR dosimetry. The choice of the settings that we have made (those providing about 52% of the maximal EPR intensity for each compound) is a conservative one. However, for alanine (6.3 mW and 0.99 mT) they are close to those prescribed in the approved EPR/Alanine dosimetry protocol (low doses) (8 mW/1.0 mT) (ASTM, 1998). The settings chosen in each case prevent excessive microwave power saturation and overmodulation. As seen from Fig. 2, the gradients of the P/MA dependence profiles are large in the vicinities of the chosen parameter values, which makes the accuracy of comparisons of materials measured on different instruments critically dependent on the accuracy of the instrument calibrations.
ARTICLE IN PRESS Li-formate Mg-formate Ca-formate
20
10
100 Alanine Li-formate Mg-formate Ca-formate
95 90 85 80
0
75 0
10
20
30
40
HVL (mm Al) Fig. 5. Dependence of the EPR signal intensities of irradiated Li-, Mg-, and Ca-formate normalized to the corresponding signal intensities of alanine on the photon energy expressed as HVL in aluminum. Individual microwave power and modulation amplitude settings were used for each compound (Table 2). See Table 1 for the characteristics of the corresponding beams.
10 (µen/ρ)x/(µen/ρ)H2O
187
105 EPR intensity (%)
EPR intensity (rel. alanine)
T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188
Alanine Li-formate Mg-formate Ca-formate
1 10
100 Photon energy (keV)
1000
Fig. 6. Dependence of the mass energy-absorption coefficients relative to water on the photon energy.
4.2. Li-formate Water equivalence in terms of attenuation and scattering of ionizing irradiation is an important criterion for a clinically acceptable dosimeter material. The response of Li-formate to the absorbed dose (to alanine) is only modestly dependent on the photon energy (Fig. 5). This is consistent with the calculated effective atomic number and the mass energy-absorption coefficients (Fig. 6) and shows that Li-formate is, in fact, closer to water in terms of the absorption properties than alanine. Li-formate can form strong pellets even without a binding material. This is advantageous to a dosimeter material, as it increases the sensitivity. The high resistance of its signal to fading is also very promising. However, it was observed in these initial pilot experi-
0
1
2 3 4 5 6 Time after irradiation (days)
7
Fig. 7. Dependence of the peak-to-peak intensities of the EPR spectra of alanine, Li-, Mg-, and Ca-formate on time after irradiation to 10 Gy (220 kV X-rays). Individual microwave power and modulation amplitude settings were used for each compound (Table 2). The initial (30–45 min after irradiation) spectrum intensity for each compound is taken for 100%.
ments that the actual fading characteristics may depend on one or several of the environmental factors like temperature, humidity and illumination during the pellet production and irradiation. In addition, the magnitude of the delivered dose may be a factor affecting the fading behavior. These uncertainties need to be investigated systematically. It is interesting to note that 7Li is the dominant isotope of Li present in the commercial samples of Liformate (92.5%), whereas 6Li exhibits a relative abundancy of 7.5%. The interaction between the unpaired electron and the 7Li nucleus is sufficiently weak not to become resolved at any power or modulation amplitude used, resulting in a generally wider line (1.5 mT). 7Li has a nuclear spin of 3/2. 6Li has a nuclear spin of 1, and a nuclear g-factor ðgN Þ that is 2.64 times smaller than that of 7Li. Thus, the interaction with the 6 Li nucleus should result in a spectral extent that is about 25% of that of 7Li, if the line width is determined primarily by one Li nuclear interaction. This might increase the relative sensitivity of 6Li-formate by a factor of up to 16 (as measured using the first-derivative peakto-peak intensity), all other factors being equal. However, the CO 2 g-anisotropy will probably become resolved so that the effective gain, as compared with 7 Li, will certainly be smaller than this. Still, the prospective sensitivity enhancement by using the Liformate nearly 100% enriched in 6Li should be elucidated. Deuterated samples may further enhance the sensitivity of Li-formate (Olsson et al., 2000). It can also be noted that the well-known sensitivity of 6Li to neutrons may make 6Li-formate a potentially useful material in the EPR-based neutron- or combined neutron/gamma dosimetry. Commercial Li-formate typically has radiation sensitivity 6–7 times higher than alanine and does not exhibit
ARTICLE IN PRESS 188
T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188
a zero-dose signal. Thus, this material has the potential of being useful in precise measurements of doses down to or even below the 0.1 Gy level. This makes it interesting for clinical dosimetry. 4.3. Mg-, Ca-formate Various less favorable properties, like rapid fading, a zero-dose signal (Mg-formate), and poor water equivalence (Ca-formate) rendered Mg-formate and Ca-formate generally unsuitable as general dosimeter materials. However, Ca-formate has some interesting EPR dosimetric properties. Due to the poor water equivalence, Ca-formate is 27 times more sensitive to low-energy photons (60 and 100 kV X-rays) than alanine. Provided that the radiation quality of the beam (photon energy or HVL-value) has been well-characterized, one can take advantage of this strong signal in special applications.
Acknowledgements We are grateful to Dr. Dag Rune Olsen and The Norwegian Radium Hospital for the use of the Mobaltron 80 therapy unit.
References ASTM, 1998. Standard Practice for Use of the Alanine-EPR Dosimetry system. Standard E1607-96, Annual Books of American Standards for Testing Materials, Vol. 12.02. 798–802. Bogushevich, S.E., Ugolev, I.I., 2000. Inorganic EPR dosimeter for medical radiology. Appl. Radiat. Isot. 52, 1217–1219. Bradshaw, W.W., Cadena, D.G., Crawford, G.W., Spetzler, H.A.W., 1962. The use of alanine as a solid dosimeter. Radiat. Res. 17, 11–21. Callens, F., Moens, P., Verbeek, R., 1995. An EPR study of intact and powdered human tooth enamel dried at 400 C. Calcif. Tissue Int. 63, 409–415. Callens, F., vanLaere, K., Mondelaers, W., Matthys, P., Boesman, E., 1996. A study of the composite character of the ESR spectrum of alanine. Appl. Radiat. Isot. 47, 1241–1250. Callens, F., Vanhaelewyn, G., Matthys, P., Boesman, E., 1998. EPR of carbonate derived radicals: application in dosimetry, dating and detection of irradiated food. Appl. Magn. Reson. 14, 235–254. Gustafsson, H., Olsson, S., Lund, A., Lund, E., 2003. Ammonium formate, a compound for sensitive EPR dosimetry. Radiat. Res., submitted for publication. Hassan, G.M., Ikeya, M., 2000. Metal ion-organic compound for high sensitive ESR dosimetry. Appl. Radiat. Isot. 52, 1247–1254. Heydari, M.Z., Malinen, E., Hole, E.O., Sagstuen, E., 2002. Alanine radicals 2. The composite polycrystalline alanine
EPR spectrum studied by ENDOR, thermal annealing, and spectrum simulations. J. Phys. Chem. A 106, 8971–8977. IAEA, 1987. Absorbed dose determination in photon and electron beams. Report TRS 277. IAEA, 2000. Absorbed dose determination in external beam radiotherapy: an international code of practice for dosimetry based on standards of absorbed dose to water. Report TRS 398. Ikeya, M., Hassan, G.M., Sasaoka, H., Kinoshita, Y., Takaki, T., Yamanaka, C., 2000. Strategy for finding ESR dosimeter materials. Appl. Radiat. Isot. 52, 1209–1215. Lund, A., Olsson, S., Bonora, M., Lund, E., Gustafsson, H., 2002. New materials for ESR dosimetry. Spectrochim Acta A 58, 1301–1311. Malinen, E., Heydari, M.Z., Sagstuen, E., Hole, E.O., 2003. Alanine radicals. Properties of the components contributing to the EPR spectrum of X-irradiated alanine dosimeters. Radiat. Res. 159, 23–32. Mehta, K., Girzikowsky, R., 1995. Reference dosimeter system of the IAEA. Radiat. Phys. Chem. 46, 1247–1250. Mehta, K., Girzikowsky, R., 2000. IAEA high-dose intercomparison in 60Co field. Appl. Radiat. Isot. 52, 1179–1184. Murali, S., Natarajan, V., Venkataramani, R., Pushparaja, Sastry, M.D., 2001. ESR dosimetry using inorganic materials: a case study of Li2CO3 and CaSO4:Dy as prospective dosimeters. Appl. Radiat. Isot. 55, 253–258. Nette, H.P., Onori, S., Fattibene, P., Regulla, D.F., Wieser, A., 1993. Coordinated research efforts for establishing an international radiotherapy dose intercomparison service based on the alanine/ESR system. Appl. Radiat. Isot. 44, 7–12. Nowotny, R., 1998. XMuDat: Photon Attenuation Data (v.1.0.1). University of Vienna, Austria. Olsson, S.K., Lund, E., Lund, A., 2000. Development of ammonium tartrate as an ESR dosimeter material for clinical purposes. Appl. Radiat. Isot. 52, 1235–1241. Orsega, F.O., Corvaja, C., 1977. Electron spin resonance of girradiated single crystals of calcium formate. J. Mol. Struct. 40, 211–214. Ovenall, D.W., Whiffen, D.H., 1961. Electron spin resonance and structure of the CO 2 radical ion. Mol. Phys. 4, 135–144. Regulla, D.F., Deffner, U., 1982. Dosimetry by ESR spectroscopy of alanine. Appl. Radiat. Isot. 33, 1101–1114. Sagstuen, E., Hole, E.O., Haugedal, S.R., Nelson, W.H., 1997. Alanine radicals: structure determination by EPR and ENDOR of single crystals X-irradiated at 295 K. J. Phys. Chem. A 101, 9763–9772. Schaeken, B., Scalliet, P., 1996. One year of experience with alanine dosimetry in radiotherapy. Appl. Radiat. Isot. 47, 1177–1182. Supe, A.A., Zubarev, V.E., Bugaenko, L.T., 1986. Radiationchemical yields of CO 2 radical in the gamma-irradiated formiates of the alkali- and alkali-earth metals. Latv. PSR Zinat. Akad. Vestis 4, 433–438. Vanhaelewyn, G.C.A.M., Mondelaers, W.K.P.G., Callens, F.J., 1999. Effect of temperature on the electron paramagnetic resonance spectrum of irradiated alanine. Radiat. Res. 151, 590–594.
Applied Radiation and Isotopes 59 (2003) 181–188
EPR dosimetric properties of formates Tor Arne Vestada, Eirik Malinena, Anders Lundb, Eli Olaug Holea, Einar Sagstuena,* a
Department of Physics, University of Oslo, P.O. Box 1048 Blindern, N-0316 Oslo, Norway b Chemical Physics Laboratory, IFM, Linkoping University, S-581 83 Linkoping, Sweden . . Received 2 April 2003; received in revised form 7 May 2003; accepted 22 June 2003
Abstract As a part of a program to develop an electron paramagnetic resonance (EPR) dosimeter suited for clinical use (doses in the cGy range), polycrystalline samples of lithium formate monohydrate (HCO2Li H2O), magnesium formate dihydrate (C2H2O4Mg 2H2O), and calcium formate (C2H2O4Ca) have been examined. l-Alanine was included for comparison and reference. Samples were irradiated with 60Co g-rays and 60–220 kV X-rays. The dosimeter response was assessed using the peak-to-peak amplitude of the first-derivative EPR spectrum. Dose–response curves for the 60Co g-irradiated samples were constructed, and the dependences of the response on the photon energy, microwave power, and modulation amplitude were studied. Stability of the irradiation products upon storage (signal fading) was also investigated. Lithium formate monohydrate is by far the best candidate of the tested formates, suitable for measuring doses down to approximately 0.1 Gy. Lithium formate monohydrate is more sensitive than alanine by a factor of 5.6– 6.8 in the tested photon energy range, it exhibits no zero-dose signal and shows a linear dose response in the dose range from 0.2 to 1000 Gy. Its EPR signal was found unchanged in shape and intensity 1 week after irradiation to 10 Gy. Various less favorable properties rendered the other formates generally unsuitable, although calcium formate exhibits some interesting EPR dosimetric properties. r 2003 Elsevier Ltd. All rights reserved. Keywords: EPR; EPR dosimetry; Lithium formate; Formates; Dosimetry; Therapeutic range
1. Introduction Electron paramagnetic resonance (EPR) dosimetry with l-alanine has become a standard technique at the IAEA for measurements of high doses and for transfer dosimetry (Nette et al., 1993; Mehta and Girzikowsky, 1995, 2000). However, even with several decades of EPR studies on alanine (Bradshaw et al., 1962; Regulla and Deffner, 1982), the basic properties of alanine are still under investigation (Callens et al., 1996; Sagstuen et al., 1997; Vanhaelewyn et al., 1999; Heydari et al., 2002; Malinen et al., 2003). In addition, searches for alternatives to alanine as a dosimetric material are *Corresponding author. Tel.: +47-228-55653; fax: +47-22855671. E-mail address: [email protected] (E. Sagstuen).
actively pursued (Bogushevich and Ugolev, 2000; Hassan and Ikeya, 2000; Ikeya et al., 2000; Olsson et al., 2000; Murali et al., 2001; Lund et al., 2002). An important motivation for such searches is that the sensitivity of alanine dosimetry is too low to make this technique a realistic alternative to the dosimetry systems currently in use in clinical practice, where doses down to 0.1 Gy should be easily detectable. Radiation-induced paramagnetic defects in formates (salts of formic acid, HCOOH) have been proposed as dose indicators for application in EPR dosimetry (Lund et al., 2002). In the present study, which is a part of a program to develop an EPR dosimeter for doses in the cGy range, polycrystalline samples of lithium formate monohydrate (HCO2Li H2O), magnesium formate dihydrate (C2H2O4Mg 2H2O), and calcium formate (C2H2O4Ca) have been examined. Pellets of the samples
0969-8043/$ - see front matter r 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0969-8043(03)00165-9
ARTICLE IN PRESS 182
T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188
were irradiated using 60Co g-rays and X-rays with energies between 60 and 220 kV, and the dependences of the EPR spectra of the various formates on microwave power and modulation amplitude were studied. The dose–response curves for the 60Co g-irradiated samples were constructed in the 0.5–1000 Gy dose range (lithium formate monohydrate was additionally examined in the dose range from 0.2 to 1.0 Gy). Furthermore, for each of the studied formates, the dependence of the EPR spectrum intensity on photon energy was investigated using the intensity of the alanine spectrum as a reference, and compared with the corresponding calculated mass energy-absorption coefficients. Finally, the stability of the irradiation products upon storage (signal fading) was studied. Pilot experiments using sodium formate (CHO2Na), potassium formate (CHO2K), ammonium formate (CH5NO2), and cesium formate (CHO2Cs) were also performed. Potassium formate was not examined further, as it was realized that utilizable pellets could not be made because of the hygroscopic properties of this material. In the case of sodium formate, a relatively large hyperfine coupling (hfc) from sodium was observed (B0.8 mT), which is unfavorable for EPR dosimetry, and the sensitivity was low as compared with the other formates tested. Ammonium formate is, however, a promising material for sensitive EPR dosimetry (Gustafsson et al., 2003). Cesium formate gave no detectable EPR signal after irradiation with Xrays (100 kV) to the dose of 100 Gy, apparently in contradiction to the previous results (Supe et al., 1986). The main signal observed in the spectra of irradiated formates is attributed to CO 2 radicals (Ovenall and Whiffen, 1961; Orsega and Corvaja, 1977; Supe et al., 1986; Lund et al., 2002). The EPR spectrum of CO 2 radicals is single-lined, making the polycrystalline spectrum from the formates fairly simple, mainly dependent on the strength of the hfc interaction with the alkali metal cation and the g-tensor anisotropy. The spectrum of HCO2Li H2O pellets is a single line with the width approximately 1.5 mT. In the other investigated formates, the interaction of the unpaired electron with the cation nucleus is weaker, the g-tensor anisotropy is resolved (under non-saturating conditions and low modulation amplitudes), and the lines are, accordingly, narrower (approximately 0.25 mT).
2. Materials and methods 2.1. Samples Polycrystalline lithium formate monohydrate (HCO2Li H2O) was obtained from Sigma-Aldrich; magnesium formate dihydrate (C2H2MgO4 2H2O) and calcium formate (C2H2CaO4) were obtained from Fluka. These
materials are hereafter referred to as Li-Mg- and Caformate. Samples were made in the form of cylindrical pellets (D ¼ 5 mm, h ¼ 471 mm) with a manual pellet press (10-kN force was applied for 1 min). Alanine pellets were obtained from Bruker BioSpin (Bruker Bronze lot No. 587895/1). The polyethylene content of the alanine pellets (20% by weight) was accounted for in the analysis of the signal intensities. During irradiation and storing, the pellets were kept in cylindrical polyethylene vials (thickness 1 mm, inner diameter 6 mm), arranged in groups three by three. The temperature and relative humidity in the laboratory during preparation, storing, and EPR measurements were 2572 C and 1572%, respectively. The pellets were stored in darkness and exposed to luminescent light only when transferred from the storage area to a radiation source or the EPR spectrometer.
2.2. Radiation The X-ray unit used was a Pantak HF225 generator, with a maximum voltage and power of 225 kV and 3.2 kW. The inherent filtration was 1 mm beryllium, and two primary filters were available, namely, 1.5 mm aluminum and 0.5 mm copper. The used combinations of voltages and filters are listed in Table 1, along with the corresponding half value layers (HVLs) in aluminum. HVL is defined as the thickness of an absorber that reduces the air kerma of an X-ray beam by 50% (IAEA, 1987, 2000). During the X-ray irradiation, vials with pellets were sitting on a 10-mm-thick PVC slab about 0.5 m away from the source. The 60Co g-irradiation was performed with a Mobaltron 80 therapy unit (TEM Instruments Limited). The vials with the pellets were located 0.80 m far from the source with a 5-mm buildup of solid water and over 100 mm of solid water as a backscatter material. In both X-ray and 60Co irradiations, electronic equilibria were achieved. All the quoted radiation doses are expressed as doses to alanine, since alanine is chosen as the reference medium in this work. For theoretical estimates of the energy response, the mass energy-absorption coefficients and the effective
Table 1 Radiation beams with the filters used and the corresponding HVL employed in this study Beam
Filter
HVL (mm Al)
60 kV 100 kV 220 kV 60 Co g-rays
1.5 mm Al 1.5 mm Al 1.5 mm Al+0.5 mm Cu —
1.6 3.0 15 42.3
ARTICLE IN PRESS T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188
atomic numbers of the samples were calculated using program XMuDat (Nowotny, 1998). 2.3. EPR measurements EPR spectra were recorded at room temperature using a Bruker ESP300E spectrometer equipped with a standard X-band bridge and a dual rectangular cavity (TE104). Upon measurements, the pellet was placed in an open quartz tube with an inner diameter of 5.25 mm, and a sample support system was used to ensure identical and reproducible positions of all samples in the cavity. A Mn2+/MgO reference sample (JEOL Co.) was placed into the second cavity, and the EPR spectrum of Mn2+ was recorded immediately after each dosimeter spectrum registration in order to correct the spectrum intensity for spectrometer instabilities. In all measurements except those in the fading experiments (see below), the EPR signal from the reference sample was recorded using a microwave power 5 mW and a modulation amplitude 0.15 mT. All reference sample measurements showed a relative standard deviation of less than 1.2%. The term ‘EPR intensity’ corresponds to the peak-to-peak amplitude of the first-derivative curve of the EPR spectrum. In the studies of the signal intensity dependences on microwave power ðPÞ and modulation amplitude (MA), one sample of each material, irradiated with 60Co g-rays to 10 Gy, was used. The microwave power was being decreased from 202 to 1 mW (0–23 dB) in steps of 1 dB, and the EPR spectrum was recorded at each step. This stepwise power change was repeated for 20 modulation amplitude settings increasing from 0.1 to 3.1 mT. Thus, 480 measurements in total have characterized the P=MA dependence profile for each material (only selected data are presented in the figures of this paper). Based on the P=MA dependence profiles, individual microwave power and modulation amplitude settings were chosen
Table 2 The individual microwave power ðPÞ and modulation amplitude (MA) settings employed in this study with the resulted EPR intensities (expressed in percent of the maximum intensity achievable in the ranges 1–202 mW and 0.1–3.1 mT) and the time between the sample irradiation and the spectrum registration
Alanine Li-formate Mg-formate Ca-formate
183
Table 3 Average relative standard deviations of the signals from replicate pellets of various dosimetric materials irradiated to the same dose Dose range (Gy)
Alanine
Li-format
Mgformate
Caformate
1.0–10 0.2–1.0
1.9% —
1.2% 4.3%
2.7% —
2.7% —
for each material (Table 2) that produce signals equal to 5271% of its maximal value (marked in Fig. 2) in the P/MA ranges investigated. This approach makes it possible to compare performance characteristics of different dosimetric materials accurately. The measurements related to the spectrum shape (60Co, 100 Gy), energy response (10 Gy), and dose response (60Co) were performed 64–70 h after irradiation. Each data point in Figs. 3 and 5 represents the average of the signals of three similar samples irradiated simultaneously; the error bars (barely visible) represent one standard deviation of the mean. Additional dose response measurements for subgray doses were performed for Li-formate, with five samples irradiated simultaneously for each dose point (Fig. 4). Average relative standard deviations for the various dosimeters are listed in Table 3. In the study of fading, one sample of each material irradiated to 10 Gy with 220 kV X-rays was used. The temperature during irradiation was 295 K. Its signal measurements were made at equal time intervals (ranged between 9 and 40 min for different materials) continuously for 6–7 days, starting 30–45 min after the end of the irradiation. The room temperature and humidity did not vary significantly during the measurement sessions. In the Mg-formate and alanine fading experiments, the reference sample signals were recorded at the 10 mW power and the 0.47 mT modulation amplitude, in contrast to the parameters specified above. No systematic trends in the reference sample amplitudes were detected that would indicate, e.g., a significant accumulation of moisture in the sample pellets during the timeextended measurement series.
3. Results
Power (mW)
MA (mT)
EPR intensity (%)
Time (h)
3.1. EPR spectra of
6.33 39.9 20.0 15.9
0.990 1.25 0.990 0.990
51.8 51.2 52.0 52.8
64 66 68 70
EPR spectra of the pellets of alanine, Li-, Mg- and Ca-formates irradiated to 100 Gy (60Co g-rays) are shown in Fig. 1. The alanine spectrum consists of the contributions from at least three different radical species (Sagstuen et al., 1997), while the relatively sharp
60
Co g-irradiated samples
ARTICLE IN PRESS T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188
184
3.2. Dependence of the EPR spectra on microwave power and modulation amplitude Alanine 1.0
Li-formate 5.6
The power and modulation amplitude dependence profiles for alanine, Li-, Mg-, and Ca-formate are shown in Fig. 2. The pellets were irradiated to 10 Gy using 60Co g-rays. The alanine profile shows a clear maximum at 50 mW and 1.6 mT. The profiles of the formates do not feature any well-defined maxima. The highest intensities, however, were achieved at the parameter values listed in Table 4. 3.3. Dose–response curves for
Mg-formate 4.2
Ca-formate 4.3 342
344
346
348
350
352
B-field (mT) Fig. 1. First-derivative EPR spectra of pellets of alanine, Li-, Mg-, and Ca-formate irradiated to 100 Gy with 60Co g-rays. The upper spectrum for each compound was recorded at P ¼ 4:0 mW and MA=0.1 mT, the lower one was recorded at the individual power and modulation amplitude settings listed in Table 2. All the spectra are normalized to the intensity of the alanine spectrum, and the normalization factors for the lower spectra are quoted under each curve. The g value at the center of each spectrum is 2.0023.
resonance from the irradiated formates has been attributed to CO 2 radicals (Ovenall and Whiffen, 1961; Orsega and Corvaja, 1977; Supe et al., 1986; Lund et al., 2002). The EPR spectra of Li-, Mg-, and Caformate are different in shape due to the different nuclear spins, nuclear g-factors and isotopic abundances of the cations (7Li 92.5%, I ¼ 3=2; gN ¼ 5:792; 6Li 7.5%, I ¼ 1; gN ¼ 0:822; 25Mg 10.13%, I ¼ 5=2; gN ¼ 0:342; 43Ca 0.14%, I ¼ 7=2; gN ¼ 0:376), and due to different anisotropies. Li-formate has an apparently symmetrical 1.5-mT wide resonance line ðg ¼ 2:0005Þ: The spectrum shape is not sensitive to the power and modulation amplitude settings (Fig. 1). The spectra of Mg- and Ca-formate are very similar to each other, and they exhibit the characteristic doublet-like pattern of the powder spectrum of CO 2 radicals well known from the literature (Callens et al., 1995, 1998). High power and modulation amplitude settings mask the doublet.
60
Co g-irradiated samples
Fig. 3 shows the dose–response curves for 60Co girradiated samples in the ranges 0.5–1000 Gy (Fig. 3a, logarithmic scale) and 0.5–10 Gy (Fig. 3b, linear scale). The individual EPR settings producing 52% of the maximal EPR intensities were used (Table 2). Li- and Ca-formate show a linear dose response in the dose range from 0.5 to 1000 Gy without any zero-dose signal. Mg-formate exhibits a substantial zero-dose signal (equivalent approximately to 1 Gy), which is different from the radiation-induced signal, and the dose dependence of the measured EPR signal intensity is not linear below 2 Gy. Alanine generally exhibits a zero-dose signal as well, and its shape and intensity depend somewhat on the production procedure (Schaeken and Scalliet, 1996). Results for Li-formate in the dose range 0.2–1.0 Gy, including the EPR spectra, are presented in Fig. 4. Liformate shows a linear dose response in this dose range. Reinsertion of the same pellet irradiated to 0.4 Gy into the spectrometer cavity 10 times resulted in a statistical sample of the signal values with the relative standard deviation of 4.5% (the data are not shown). 3.4. Effect of the photon energy Fig. 5 shows the dependence of the dose responses of the studied materials on the photon energy. The intensities of the signals of each material were normalized to the intensity of the signal of alanine irradiated to the same dose with the photons of the same energy. The qualities of the used photon beams expressed as HVL in Al are listed in Table 1. The individual EPR settings producing 52% of the maximal EPR intensities were used (Table 2). It appears from Fig. 5 that Li-formate is more sensitive to the photon energy than alanine by a factor of 5.6–6.8 in the tested radiation quality range (HVL from 1.6 to 42.3 mm in Al). Furthermore, of the materials tested, Li-formate appears to be the least sensitive in its response (relative to alanine) to the photon energy. Mg-formate exceeds alanine in the sensitivity to the photon energy by a factor of 4.2–8.2.
ARTICLE IN PRESS T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188 MA (mT)
100 1.57 1.97 1.25 2.49 0.99 0.79 3.13 0.62
75 50 25
0.39
MA (mT)
Li-formate EPR intensity (rel. units)
EPR intensity (rel. units)
Alanine
185
100
3.13 2.49 1.97
75
1.57 1.25
50
0.99 0.79 0.62
25
0.39 0.20
0.20
0 0
100 150 Power (mW)
200
0
100
150
200
Power (mW) MA (mT)
Mg-formate EPR intensity (rel. units)
05
2.49 1.97 3.13 1.57 1.25
100 75
0.99 0.79
50
0.62 0.39
25
MA (mT)
Ca-formate EPR intensity (rel. units)
05
1.97 2.49 1.57 3.13 1.25
100
0.99
75
0.79 0.62
50
0.39
25 0.20
0.20
0 05
0
100
150
200
Power (mW)
05
0
100
150
200
Power (mW)
Fig. 2. Dependence of the peak-to-peak intensities of the EPR spectra of alanine, Li-, Mg-, and Ca-formate irradiated to 10 Gy (60Co) on microwave power (1–202 mW) and modulation amplitude (0.1–3.1 mT, selected data). The individual microwave power and modulation amplitude settings chosen for the experiments in this work are marked in the figure. Each profile is scaled to make the maximal intensity equal to 100.
Table 4 The EPR spectrometer settings corresponding to maximum EPR intensity achievable in the ranges 1–202 mW and 0.1–3.1 mT
Alanine Li-formate Mg-formate Ca-formate
Power (mW)
MA (mT)
EPR intensity (%)
50 202 202 202
1.6 3.1 2.0–3.1 2.0–3.1
100 100 98–100 98–100
Ca-formate is about 27 times more sensitive than alanine at low photon energies, but the relative sensitivity decreases to the 4.3-fold for 60Co g-radiation. 3.5. Mass energy-absorption coefficients The effective atomic numbers of alanine, Li-, Mg-, and Ca-formate are 6.78, 7.31, 8.64, and 14.5, respectively, while the corresponding characteristic for water is 7.51 (Nowotny, 1998). The calculated dependences of the mass energy-absorption coefficients relative to water on photon energy in the range from 10 keV to 10 MeV
are shown in Fig. 6. The dependence is the weakest for Li-formate (the coefficients change between 0.84 and 0.93), while the variation ranges for alanine, Mg- and Ca-formate are 0.68–0.97, 1.50–0.93, and 7.85–0.91, respectively. 3.6. Stability of the irradiation products (signal fading characteristics) Fig. 7 shows the changes in the intensities of the EPR spectra of alanine, Li-, Mg-, and Ca-formate in time after irradiation to 10 Gy with 220 kV X-rays. The intensities at the start of the measurements (30–45 min after end of irradiation) are set to 100%. The signals have been corrected for variations in the Mn2+ reference signal. The large scatter in the results for alanine as compared with those for the formates is due to the relatively weak signals obtainable from alanine at this dose. The EPR signal of Li-formate remains unchanged in shape and intensity 1 week after irradiation to 10 Gy. The signal of Mg-formate fades to approximately 80% of the initial intensity in 7 days. The signal of Caformate increases by approximately 3% in the first few hours, but subsequently fades to approximately 75% of the initial intensity in 7 days.
ARTICLE IN PRESS T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188 1000
100
100 10
EPR intensity (arb. units)
Alanine Li-formate Mg-formate Ca-formate Regression
1 0.1 1
(a) EPR intensity (arb. units)
Li-formate Regression
80 60 40 20 0
0.1
10 100 Dose (Gy)
1000
Alanine Li-formate Mg-formate Ca-formate Regression
8 6
0 2
4 6 Dose (Gy)
0.4 0.6 Dose (Gy)
8
10
Fig. 3. Dose dependence of the peak-to-peak intensities of the EPR spectra of alanine, Li-, Mg-, and Ca-formate irradiated with 60Co g-rays. Individual microwave power and modulation amplitude settings were used for each compound (Table 2).
b)
0.8
1.0
0 Gy 0.2 Gy 0.4 Gy 0.6 Gy 0.8 Gy 1.0 Gy
Li-formate
2
0
0.2
40
4
(b)
0.0 (a)
EPR intensity (arb. units)
EPR intensity (arb. units)
186
20
0
-20
-40
347
348
349
350
B-field (mT)
4. Discussion 4.1. General The different materials tested in this study all exhibit linear dose responses in the dose range tested, except of Mg-formate, whose significant zero-dose signal results in a deviation from linearity at the lowest doses. The spectrum of Li-formate has the most convenient shape from the viewpoint of EPR dosimetry, as the spectrum appears less sensitive to the microwave power and modulation amplitudes in the tested ranges. The effect of the atomic number on the absorption of photons with energies below 250 keV is demonstrated experimentally and theoretically in Figs. 4 and 5. The response of alanine itself is clearly dependent on the photon energy (Fig. 6) due to a relatively low effective atomic number as compared with water, which makes the use of this material in soft X-ray dosimetry less straightforward. In the present work, individual power and modulation amplitude (P/MA) settings were used for each material, they are listed in Table 2. These settings provide equal EPR intensities for each material relative to the maximal achievable values, which were determined from the
Fig. 4. (a) Dependence of the peak-to-peak intensity of the EPR spectrum of Li-formate on the absorbed dose to alanine in the sub-gray range. 60Co g-rays. Five replicate samples for each dose. (b) Examples of the corresponding first-derivative EPR spectra. Microwave power 39.9 mW, modulation amplitude 1.25 mT, 10 scans, total acquisition time 7 min. The background signal from the cavity has been subtracted from the spectra.
profiles shown in Fig. 2. This appears to be a necessary approach when different materials need to be compared with each other in terms of their performance characteristics in EPR dosimetry. The choice of the settings that we have made (those providing about 52% of the maximal EPR intensity for each compound) is a conservative one. However, for alanine (6.3 mW and 0.99 mT) they are close to those prescribed in the approved EPR/Alanine dosimetry protocol (low doses) (8 mW/1.0 mT) (ASTM, 1998). The settings chosen in each case prevent excessive microwave power saturation and overmodulation. As seen from Fig. 2, the gradients of the P/MA dependence profiles are large in the vicinities of the chosen parameter values, which makes the accuracy of comparisons of materials measured on different instruments critically dependent on the accuracy of the instrument calibrations.
ARTICLE IN PRESS Li-formate Mg-formate Ca-formate
20
10
100 Alanine Li-formate Mg-formate Ca-formate
95 90 85 80
0
75 0
10
20
30
40
HVL (mm Al) Fig. 5. Dependence of the EPR signal intensities of irradiated Li-, Mg-, and Ca-formate normalized to the corresponding signal intensities of alanine on the photon energy expressed as HVL in aluminum. Individual microwave power and modulation amplitude settings were used for each compound (Table 2). See Table 1 for the characteristics of the corresponding beams.
10 (µen/ρ)x/(µen/ρ)H2O
187
105 EPR intensity (%)
EPR intensity (rel. alanine)
T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188
Alanine Li-formate Mg-formate Ca-formate
1 10
100 Photon energy (keV)
1000
Fig. 6. Dependence of the mass energy-absorption coefficients relative to water on the photon energy.
4.2. Li-formate Water equivalence in terms of attenuation and scattering of ionizing irradiation is an important criterion for a clinically acceptable dosimeter material. The response of Li-formate to the absorbed dose (to alanine) is only modestly dependent on the photon energy (Fig. 5). This is consistent with the calculated effective atomic number and the mass energy-absorption coefficients (Fig. 6) and shows that Li-formate is, in fact, closer to water in terms of the absorption properties than alanine. Li-formate can form strong pellets even without a binding material. This is advantageous to a dosimeter material, as it increases the sensitivity. The high resistance of its signal to fading is also very promising. However, it was observed in these initial pilot experi-
0
1
2 3 4 5 6 Time after irradiation (days)
7
Fig. 7. Dependence of the peak-to-peak intensities of the EPR spectra of alanine, Li-, Mg-, and Ca-formate on time after irradiation to 10 Gy (220 kV X-rays). Individual microwave power and modulation amplitude settings were used for each compound (Table 2). The initial (30–45 min after irradiation) spectrum intensity for each compound is taken for 100%.
ments that the actual fading characteristics may depend on one or several of the environmental factors like temperature, humidity and illumination during the pellet production and irradiation. In addition, the magnitude of the delivered dose may be a factor affecting the fading behavior. These uncertainties need to be investigated systematically. It is interesting to note that 7Li is the dominant isotope of Li present in the commercial samples of Liformate (92.5%), whereas 6Li exhibits a relative abundancy of 7.5%. The interaction between the unpaired electron and the 7Li nucleus is sufficiently weak not to become resolved at any power or modulation amplitude used, resulting in a generally wider line (1.5 mT). 7Li has a nuclear spin of 3/2. 6Li has a nuclear spin of 1, and a nuclear g-factor ðgN Þ that is 2.64 times smaller than that of 7Li. Thus, the interaction with the 6 Li nucleus should result in a spectral extent that is about 25% of that of 7Li, if the line width is determined primarily by one Li nuclear interaction. This might increase the relative sensitivity of 6Li-formate by a factor of up to 16 (as measured using the first-derivative peakto-peak intensity), all other factors being equal. However, the CO 2 g-anisotropy will probably become resolved so that the effective gain, as compared with 7 Li, will certainly be smaller than this. Still, the prospective sensitivity enhancement by using the Liformate nearly 100% enriched in 6Li should be elucidated. Deuterated samples may further enhance the sensitivity of Li-formate (Olsson et al., 2000). It can also be noted that the well-known sensitivity of 6Li to neutrons may make 6Li-formate a potentially useful material in the EPR-based neutron- or combined neutron/gamma dosimetry. Commercial Li-formate typically has radiation sensitivity 6–7 times higher than alanine and does not exhibit
ARTICLE IN PRESS 188
T.A. Vestad et al. / Applied Radiation and Isotopes 59 (2003) 181–188
a zero-dose signal. Thus, this material has the potential of being useful in precise measurements of doses down to or even below the 0.1 Gy level. This makes it interesting for clinical dosimetry. 4.3. Mg-, Ca-formate Various less favorable properties, like rapid fading, a zero-dose signal (Mg-formate), and poor water equivalence (Ca-formate) rendered Mg-formate and Ca-formate generally unsuitable as general dosimeter materials. However, Ca-formate has some interesting EPR dosimetric properties. Due to the poor water equivalence, Ca-formate is 27 times more sensitive to low-energy photons (60 and 100 kV X-rays) than alanine. Provided that the radiation quality of the beam (photon energy or HVL-value) has been well-characterized, one can take advantage of this strong signal in special applications.
Acknowledgements We are grateful to Dr. Dag Rune Olsen and The Norwegian Radium Hospital for the use of the Mobaltron 80 therapy unit.
References ASTM, 1998. Standard Practice for Use of the Alanine-EPR Dosimetry system. Standard E1607-96, Annual Books of American Standards for Testing Materials, Vol. 12.02. 798–802. Bogushevich, S.E., Ugolev, I.I., 2000. Inorganic EPR dosimeter for medical radiology. Appl. Radiat. Isot. 52, 1217–1219. Bradshaw, W.W., Cadena, D.G., Crawford, G.W., Spetzler, H.A.W., 1962. The use of alanine as a solid dosimeter. Radiat. Res. 17, 11–21. Callens, F., Moens, P., Verbeek, R., 1995. An EPR study of intact and powdered human tooth enamel dried at 400 C. Calcif. Tissue Int. 63, 409–415. Callens, F., vanLaere, K., Mondelaers, W., Matthys, P., Boesman, E., 1996. A study of the composite character of the ESR spectrum of alanine. Appl. Radiat. Isot. 47, 1241–1250. Callens, F., Vanhaelewyn, G., Matthys, P., Boesman, E., 1998. EPR of carbonate derived radicals: application in dosimetry, dating and detection of irradiated food. Appl. Magn. Reson. 14, 235–254. Gustafsson, H., Olsson, S., Lund, A., Lund, E., 2003. Ammonium formate, a compound for sensitive EPR dosimetry. Radiat. Res., submitted for publication. Hassan, G.M., Ikeya, M., 2000. Metal ion-organic compound for high sensitive ESR dosimetry. Appl. Radiat. Isot. 52, 1247–1254. Heydari, M.Z., Malinen, E., Hole, E.O., Sagstuen, E., 2002. Alanine radicals 2. The composite polycrystalline alanine
EPR spectrum studied by ENDOR, thermal annealing, and spectrum simulations. J. Phys. Chem. A 106, 8971–8977. IAEA, 1987. Absorbed dose determination in photon and electron beams. Report TRS 277. IAEA, 2000. Absorbed dose determination in external beam radiotherapy: an international code of practice for dosimetry based on standards of absorbed dose to water. Report TRS 398. Ikeya, M., Hassan, G.M., Sasaoka, H., Kinoshita, Y., Takaki, T., Yamanaka, C., 2000. Strategy for finding ESR dosimeter materials. Appl. Radiat. Isot. 52, 1209–1215. Lund, A., Olsson, S., Bonora, M., Lund, E., Gustafsson, H., 2002. New materials for ESR dosimetry. Spectrochim Acta A 58, 1301–1311. Malinen, E., Heydari, M.Z., Sagstuen, E., Hole, E.O., 2003. Alanine radicals. Properties of the components contributing to the EPR spectrum of X-irradiated alanine dosimeters. Radiat. Res. 159, 23–32. Mehta, K., Girzikowsky, R., 1995. Reference dosimeter system of the IAEA. Radiat. Phys. Chem. 46, 1247–1250. Mehta, K., Girzikowsky, R., 2000. IAEA high-dose intercomparison in 60Co field. Appl. Radiat. Isot. 52, 1179–1184. Murali, S., Natarajan, V., Venkataramani, R., Pushparaja, Sastry, M.D., 2001. ESR dosimetry using inorganic materials: a case study of Li2CO3 and CaSO4:Dy as prospective dosimeters. Appl. Radiat. Isot. 55, 253–258. Nette, H.P., Onori, S., Fattibene, P., Regulla, D.F., Wieser, A., 1993. Coordinated research efforts for establishing an international radiotherapy dose intercomparison service based on the alanine/ESR system. Appl. Radiat. Isot. 44, 7–12. Nowotny, R., 1998. XMuDat: Photon Attenuation Data (v.1.0.1). University of Vienna, Austria. Olsson, S.K., Lund, E., Lund, A., 2000. Development of ammonium tartrate as an ESR dosimeter material for clinical purposes. Appl. Radiat. Isot. 52, 1235–1241. Orsega, F.O., Corvaja, C., 1977. Electron spin resonance of girradiated single crystals of calcium formate. J. Mol. Struct. 40, 211–214. Ovenall, D.W., Whiffen, D.H., 1961. Electron spin resonance and structure of the CO 2 radical ion. Mol. Phys. 4, 135–144. Regulla, D.F., Deffner, U., 1982. Dosimetry by ESR spectroscopy of alanine. Appl. Radiat. Isot. 33, 1101–1114. Sagstuen, E., Hole, E.O., Haugedal, S.R., Nelson, W.H., 1997. Alanine radicals: structure determination by EPR and ENDOR of single crystals X-irradiated at 295 K. J. Phys. Chem. A 101, 9763–9772. Schaeken, B., Scalliet, P., 1996. One year of experience with alanine dosimetry in radiotherapy. Appl. Radiat. Isot. 47, 1177–1182. Supe, A.A., Zubarev, V.E., Bugaenko, L.T., 1986. Radiationchemical yields of CO 2 radical in the gamma-irradiated formiates of the alkali- and alkali-earth metals. Latv. PSR Zinat. Akad. Vestis 4, 433–438. Vanhaelewyn, G.C.A.M., Mondelaers, W.K.P.G., Callens, F.J., 1999. Effect of temperature on the electron paramagnetic resonance spectrum of irradiated alanine. Radiat. Res. 151, 590–594.