- Email: [email protected]
An example of abnormal glow curves identification in personnel thermoluminescent dosimetry
J. Biochem. Biophys. Methods 53 (2002) 117 – 122 www.elsevier.com/locate/jbbm
An example of abnormal glow curves identification in personnel thermoluminescent dosimetry V. Osorio Piniella a, H. Stadtmann a, E. Lankmayr b,* b
a Austrian Research Centre Seibersdorf, A-2444 Seibersdorf, Austria Micro- and Radiochemistry, Institute for Analytical Chemistry, Graz University of Technology, Technikerstrasse 4/P, A-8010 Graz, Austria
Received 6 August 2001; accepted 14 December 2001
Abstract The personal Dosimetry Service Seibersdorf analyses monthly a large number of thermoluminescent dosimeters (TLD). The dosimeters consist of LiF chips, and the readout is carried out with an automated Harshaw 8800 reader system. In some cases, the luminescent glow curves of the routine analysis do not have the expected form as a result of external chemical contamination, hardware problems, poor heat transfer, etc. It is therefore necessary to investigate the reasons for the irregularity of these curves. An algorithm for the investigation of the routine curves was developed. It is based on the fact that the shape of an abnormal glow curve differs from the shape of a normal one. An interesting type of abnormal glow curves in the routine service was found. Some dosimeters of a certain client, a steel industry, exhibit glow curves with an untypical shape and very high signals. In those dosimeters, a possible chemical contamination in the form of a powder was discovered, which interferes with the dosimetric signal. A quantitative analysis of that powder was made by means of inductively coupled plasma emission spectroscopy (ICP-OES) after microwave dissolution. Elements like aluminium, barium, calcium and others were found. Such elements are used in different combinations as thermoluminescent materials. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Dosimetry; Irregular glow curves; TLD-100; Thermoluminescence
1. Introduction Thermoluminescence is a process in which certain crystalline materials, in our case the detector element, store energy when they are exposed to ionising radiation. With posterior *
Corresponding author. Tel.: +43-316-873-8305; fax: +43-316-873-8304. E-mail address: [email protected] (E. Lankmayr).
0165-022X/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 0 2 2 X ( 0 2 ) 0 0 0 9 9 - 4
118
V. Osorio Piniella et al. / J. Biochem. Biophys. Methods 53 (2002) 117–122
heating to the material, a small amount of light is emitted which can be sensed and amplified by a photomultiplier tube. This emitted light is not instantaneous and depends on the detector temperature which is unique for each type of crystal material. The light to temperature relationship is called glow curve. One crystal, which exhibits pronounced and repeatable glow curve characteristics and which is widely employed therefore, is lithium fluoride. When LiF is used as crystal doped with magnesium and titanium, the resulting material is called LiF:Ti,Mg or TLD-100 [1]. The analysis of thermoluminescent dosimeters (TLD) stemming from various clients needs to be processed on a routine basis. Dose information is extracted from the integral area under the glow curve. For low dose values, one of the major sources of uncertainty in the measurement of radiation doses is the presence of background signals. This background can be induced by several disturbing effects and consists of signals that are not caused by ionising radiation. The influence of these effects is recognised in the shape of the curves because the glow curve of a radiation-exposed TLD crystal has a well-known and defined shape. The typical spurious effects in glow curves are well known, and it is possible to find many examples in the related literature. A curve can have an abnormal shape due to UV radiation, microwaves, electric sparks, etc. External contaminants such as chalk and chalk dust, normal dust, oil, hand-lotions, etc. on or absorbed in the crystal are another potential source for interferences. These effects can be diminished by cleaning the detector with alcohol. It is also important to consider the light sensitivity of the PTFE matrix and a probable oxidation effect on the surface of LiF detectors which may produce chemiluminescence [2– 7]. In this work, investigations concerning a special spurious problem of the glow curves are presented. These investigations are based on our algorithm [8] developed to identify the abnormal curves by comparing the shapes with that of a standard curve.
2. Materials and methods The dosimeters consist of TLD-100 chips (BICRON) mounted between PTFE foils on aluminium cards. For personnel monitoring, these cards are wrapped and heat-sealed in thin aluminised plastic bags. The cards are worn in plastic badges with one open window and an aluminium filter of 1000 mg/cm2. The readout of the TL detector is carried out with an automated Harshaw 8800 reader unit. In these systems, detector elements are heated with a programmed time – temperature profile, and the emitted light is sensed and amplified by a photomultiplier tube. For the chemical analysis of the contaminant, powder was collected from the surface of the dosimeters. A quantity of 80-mg powder was weighted into quartz vessels and 4 ml HNO3 and 0.5 ml HCl were added. Decomposition of the inorganic sample matrix was accomplished with a Paar Multiwave digestion apparatus (Paar Physica, Graz, Austria), and the following conditions were programmed: 1000 W, 45 min, 250 jC and 75 Pa. The liquid sample solution was analysed by inductively coupled plasma emission spectroscopy (ICP-OES) with a Perkin Elmer Optima 3000 system (Perkin Elmer, Norwalk, CT, USA).
V. Osorio Piniella et al. / J. Biochem. Biophys. Methods 53 (2002) 117–122
119
Fig. 1. Algorithm to identify abnormal glow curves in the routine service.
2.1. Algorithm development Due to the large number of dosimeters, which are monthly analysed in a Personnel Dosimetry Service, spurious effects can also be observed. It is necessary therefore to identify the abnormal curves in the cycle of routine analysis. The purpose of the ongoing study was therefore the development and testing of a suitable new algorithm for an automated recognition of shape irregularity. It is based on the fact that the shape of an abnormal glow curve differs from the shape of a normal one. In order to avoid the influence of the magnitude of the dose, the glow curves are normalised with respect to the total area. The criteria necessary to differentiate regular and irregular glow curves have been defined by comparison of a set of different irregular glow curves from a monthly exposition with a set of regular glow curves measured with the same time –temperature profile as in routine. For this purpose, each curve is divided in four regions of interest (ROIs), and the differences in the areas of the ROIs and their relationships are compared for regular and irregular curves. There are a few criteria combining all four ROIs which are necessary to distinguish between normal and abnormal glow curves. Also, the maximum of the curve is used to recognise spiked glow curves. In Fig. 1, a block diagram of the algorithm is shown. The constants used for these calculations are derived from a comparison of a set of different irregular curves with 500 standard curves, and result from quotients of normalised areas of different ROIs as well as the relative positions of the peak maxima. A more detailed explanation of the algorithm and its sensitivity with respect to typical patterns of irregularity has been published previously [8]. The algorithm permits to identify irregular glow curves for a routine service, even for small doses (>70 ASv).
3. Results and discussion This algorithm was tested for its effective application in the routine service. An interesting type of abnormal glow curves in the routine service has been studied by
120
V. Osorio Piniella et al. / J. Biochem. Biophys. Methods 53 (2002) 117–122
Fig. 2. Typical routine glow curve (left) and glow curve of a dosimeter contaminated with powder (right).
Duftschmid et al. [9], and such irregular curves could also be identified automatically for routine application. Some dosimeters of a certain client, a steel industry, exhibit glow curves with an untypical form and very high signals (up to 50 mSv). After exhaustive investigations, a possible chemical contamination in the form of a powder was discovered, which interferes with the dosimetric signal and thus produces false readings. Fig. 2 shows a typical regular routine glow curve and the glow curve of a dosimeter contaminated with that powder. In order to obtain some information on the chemical composition of the material responsible for the interferences, the powder was analysed by ICP-OES. Since the sample used for the analysis ought to be liquid, sample preparation was accomplished by microwave dissolution. After dissolution, a minor residue remained, containing mainly silica and some refractory metals like wolfram, molybdenum or the like. Only the liquid solution was used for further analysis by ICP-OES. The results of the analysis are listed in Table 1. Table 1 Results of the powder analysis with ICP-AES Element
mg/g
Al Ba Ca Cr Cu Fe K Mg Mn Na Ni P Pb S Sn Sr Ti V Zn
32 0.3 115 2.5 0.15 115 7.5 88 12 14 0.2 1.5 0.3 5.5 0.6 0.4 0.5 0.12 2
V. Osorio Piniella et al. / J. Biochem. Biophys. Methods 53 (2002) 117–122
121
The qualitative and quantitative analysis of the sample solution revealed elements like aluminium, barium, calcium or magnesium, which could also produce thermoluminescent signals. These elements are used in different compounds as thermoluminescent materials (for example Al2O3, BaF2, BaSO4, CaF2, CaSO4, MgB4O7, MgO, Mg2SiO4 and MgF2). The presence of those elements in the surface contaminant can produce therefore new signals which interfere with the dosimetric signal of interest. These signals can be produced by means of excitation or processes like photo-, tribo-, chemiluminescence, etc. Additional investigations were carried out in order to identify the chemical form of those elements in the powder, and to recognise the way in which the interference occurs. It was found that the powder contains Al2O3 and CaF2, both well known for their thermoluminescent characteristics. Therefore, the reading of dosimeters contaminated with such a dust produces irregular dosimetric signals, the TL signals of Al2O3 and CaF2 and the responses of other thermoluminescent contaminants contained on the dosimeters according to environmental conditions of the client.
4. Simplified description of the method and its application A simple algorithm, which permits the identification of irregular glow curves in routine dosimetry, was developed. The basis of the algorithm is the differences in the shape of a regular and an irregular glow curve. Presently, the algorithm is being tested for its effectiveness in routine application. Different investigations were made, and an apparent problem is the existence of a chemical contaminant in the surface of some dosimeters. Now, the reasons for the irregularity have been studied and these spurious effects can be mainly eliminated by cleaning the dosimeters with methanol. Additionally, studies with the irregular curves recognised by application of the algorithm facilitate the identification of other unknown spurious effects and will thus help to eliminate interferences from chemical contaminations.
References [1] McLinlay AF. Thermoluminescence dosimetry. Medical Physics Handbooks, vol. 5. Bristol: Adam Hilger; 1981. p. 32 – 37. [2] Hoots S, Landrum V. Glow-curve analysis for verification of dose in LiF chips. Health-Physics-USA 1982;43(6):905 – 12. [3] Lawson BJ, Moleski M, Karandy R, Phillips CR, Champagne R, Winslow RJ. Harshaw/Bicron 8807 Environmental Dosimeter—Implementation Issues, TLD Users Symposium: Changes and Opportunities in Dosimetry. Las Vegas, 13 – 17 March 1995. Reprints of Presentations; 1995. [4] German V, Weinstein M, Ben-Shachar B, Pellet O. The contribution of Teflon to the thermoluminescent signal, Israel Radiation Protection Association (Israel). The Second Regional Mediterranean Congress on Radiation Protection, 20th Regional Congress of the Israel Radiation Protection Association, Tel-Aviv, 16 – 20 Nov. 1997. Program and Extended Abstracts 321 1997;39 – 42. INIS: 29-045616. [5] Weinstein M, Ben-Shachar B, German V. Determination of the contribution of thermoluminescent signals caused by disturbing effects, Israel Radiation Protection Association (Israel). The Second Regional Mediterranean Congress on Radiation Protection, 20th Regional Congress of the Israel Radiation Protection Association, Tel-Aviv, 16 – 20 Nov. 1997. Program and Extended Abstracts 321 1997;221 – 4. INIS: 29-045666.
122
V. Osorio Piniella et al. / J. Biochem. Biophys. Methods 53 (2002) 117–122
[6] Uray I. A background treatment for TLDs. Radiat Prot Dosim 1992;40(4):275 – 8. [7] German V, Weinstein M, Ben-Shachar B. Thermoluminescent signals caused by disturbing effects. Radiat Prot Dosim 1999;84(1 – 4):167 – 70. [8] Osorio Piniella V, Stadtmann H, Lankmayr E. A new algorithm to identify abnormal glow curves in TL personnel dosimetry. Radiat Prot Dosim 2001;96(1 – 3):139 – 41. [9] Duftschmid KE, Stadtmann H, Strachotinsky Ch, Winker N. A new type of spurious reading in TL personnel monitoring. Radiat Prot Dosim 1996;66(1 – 4):49 – 52.
An example of abnormal glow curves identification in personnel thermoluminescent dosimetry V. Osorio Piniella a, H. Stadtmann a, E. Lankmayr b,* b
a Austrian Research Centre Seibersdorf, A-2444 Seibersdorf, Austria Micro- and Radiochemistry, Institute for Analytical Chemistry, Graz University of Technology, Technikerstrasse 4/P, A-8010 Graz, Austria
Received 6 August 2001; accepted 14 December 2001
Abstract The personal Dosimetry Service Seibersdorf analyses monthly a large number of thermoluminescent dosimeters (TLD). The dosimeters consist of LiF chips, and the readout is carried out with an automated Harshaw 8800 reader system. In some cases, the luminescent glow curves of the routine analysis do not have the expected form as a result of external chemical contamination, hardware problems, poor heat transfer, etc. It is therefore necessary to investigate the reasons for the irregularity of these curves. An algorithm for the investigation of the routine curves was developed. It is based on the fact that the shape of an abnormal glow curve differs from the shape of a normal one. An interesting type of abnormal glow curves in the routine service was found. Some dosimeters of a certain client, a steel industry, exhibit glow curves with an untypical shape and very high signals. In those dosimeters, a possible chemical contamination in the form of a powder was discovered, which interferes with the dosimetric signal. A quantitative analysis of that powder was made by means of inductively coupled plasma emission spectroscopy (ICP-OES) after microwave dissolution. Elements like aluminium, barium, calcium and others were found. Such elements are used in different combinations as thermoluminescent materials. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Dosimetry; Irregular glow curves; TLD-100; Thermoluminescence
1. Introduction Thermoluminescence is a process in which certain crystalline materials, in our case the detector element, store energy when they are exposed to ionising radiation. With posterior *
Corresponding author. Tel.: +43-316-873-8305; fax: +43-316-873-8304. E-mail address: [email protected] (E. Lankmayr).
0165-022X/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 0 2 2 X ( 0 2 ) 0 0 0 9 9 - 4
118
V. Osorio Piniella et al. / J. Biochem. Biophys. Methods 53 (2002) 117–122
heating to the material, a small amount of light is emitted which can be sensed and amplified by a photomultiplier tube. This emitted light is not instantaneous and depends on the detector temperature which is unique for each type of crystal material. The light to temperature relationship is called glow curve. One crystal, which exhibits pronounced and repeatable glow curve characteristics and which is widely employed therefore, is lithium fluoride. When LiF is used as crystal doped with magnesium and titanium, the resulting material is called LiF:Ti,Mg or TLD-100 [1]. The analysis of thermoluminescent dosimeters (TLD) stemming from various clients needs to be processed on a routine basis. Dose information is extracted from the integral area under the glow curve. For low dose values, one of the major sources of uncertainty in the measurement of radiation doses is the presence of background signals. This background can be induced by several disturbing effects and consists of signals that are not caused by ionising radiation. The influence of these effects is recognised in the shape of the curves because the glow curve of a radiation-exposed TLD crystal has a well-known and defined shape. The typical spurious effects in glow curves are well known, and it is possible to find many examples in the related literature. A curve can have an abnormal shape due to UV radiation, microwaves, electric sparks, etc. External contaminants such as chalk and chalk dust, normal dust, oil, hand-lotions, etc. on or absorbed in the crystal are another potential source for interferences. These effects can be diminished by cleaning the detector with alcohol. It is also important to consider the light sensitivity of the PTFE matrix and a probable oxidation effect on the surface of LiF detectors which may produce chemiluminescence [2– 7]. In this work, investigations concerning a special spurious problem of the glow curves are presented. These investigations are based on our algorithm [8] developed to identify the abnormal curves by comparing the shapes with that of a standard curve.
2. Materials and methods The dosimeters consist of TLD-100 chips (BICRON) mounted between PTFE foils on aluminium cards. For personnel monitoring, these cards are wrapped and heat-sealed in thin aluminised plastic bags. The cards are worn in plastic badges with one open window and an aluminium filter of 1000 mg/cm2. The readout of the TL detector is carried out with an automated Harshaw 8800 reader unit. In these systems, detector elements are heated with a programmed time – temperature profile, and the emitted light is sensed and amplified by a photomultiplier tube. For the chemical analysis of the contaminant, powder was collected from the surface of the dosimeters. A quantity of 80-mg powder was weighted into quartz vessels and 4 ml HNO3 and 0.5 ml HCl were added. Decomposition of the inorganic sample matrix was accomplished with a Paar Multiwave digestion apparatus (Paar Physica, Graz, Austria), and the following conditions were programmed: 1000 W, 45 min, 250 jC and 75 Pa. The liquid sample solution was analysed by inductively coupled plasma emission spectroscopy (ICP-OES) with a Perkin Elmer Optima 3000 system (Perkin Elmer, Norwalk, CT, USA).
V. Osorio Piniella et al. / J. Biochem. Biophys. Methods 53 (2002) 117–122
119
Fig. 1. Algorithm to identify abnormal glow curves in the routine service.
2.1. Algorithm development Due to the large number of dosimeters, which are monthly analysed in a Personnel Dosimetry Service, spurious effects can also be observed. It is necessary therefore to identify the abnormal curves in the cycle of routine analysis. The purpose of the ongoing study was therefore the development and testing of a suitable new algorithm for an automated recognition of shape irregularity. It is based on the fact that the shape of an abnormal glow curve differs from the shape of a normal one. In order to avoid the influence of the magnitude of the dose, the glow curves are normalised with respect to the total area. The criteria necessary to differentiate regular and irregular glow curves have been defined by comparison of a set of different irregular glow curves from a monthly exposition with a set of regular glow curves measured with the same time –temperature profile as in routine. For this purpose, each curve is divided in four regions of interest (ROIs), and the differences in the areas of the ROIs and their relationships are compared for regular and irregular curves. There are a few criteria combining all four ROIs which are necessary to distinguish between normal and abnormal glow curves. Also, the maximum of the curve is used to recognise spiked glow curves. In Fig. 1, a block diagram of the algorithm is shown. The constants used for these calculations are derived from a comparison of a set of different irregular curves with 500 standard curves, and result from quotients of normalised areas of different ROIs as well as the relative positions of the peak maxima. A more detailed explanation of the algorithm and its sensitivity with respect to typical patterns of irregularity has been published previously [8]. The algorithm permits to identify irregular glow curves for a routine service, even for small doses (>70 ASv).
3. Results and discussion This algorithm was tested for its effective application in the routine service. An interesting type of abnormal glow curves in the routine service has been studied by
120
V. Osorio Piniella et al. / J. Biochem. Biophys. Methods 53 (2002) 117–122
Fig. 2. Typical routine glow curve (left) and glow curve of a dosimeter contaminated with powder (right).
Duftschmid et al. [9], and such irregular curves could also be identified automatically for routine application. Some dosimeters of a certain client, a steel industry, exhibit glow curves with an untypical form and very high signals (up to 50 mSv). After exhaustive investigations, a possible chemical contamination in the form of a powder was discovered, which interferes with the dosimetric signal and thus produces false readings. Fig. 2 shows a typical regular routine glow curve and the glow curve of a dosimeter contaminated with that powder. In order to obtain some information on the chemical composition of the material responsible for the interferences, the powder was analysed by ICP-OES. Since the sample used for the analysis ought to be liquid, sample preparation was accomplished by microwave dissolution. After dissolution, a minor residue remained, containing mainly silica and some refractory metals like wolfram, molybdenum or the like. Only the liquid solution was used for further analysis by ICP-OES. The results of the analysis are listed in Table 1. Table 1 Results of the powder analysis with ICP-AES Element
mg/g
Al Ba Ca Cr Cu Fe K Mg Mn Na Ni P Pb S Sn Sr Ti V Zn
32 0.3 115 2.5 0.15 115 7.5 88 12 14 0.2 1.5 0.3 5.5 0.6 0.4 0.5 0.12 2
V. Osorio Piniella et al. / J. Biochem. Biophys. Methods 53 (2002) 117–122
121
The qualitative and quantitative analysis of the sample solution revealed elements like aluminium, barium, calcium or magnesium, which could also produce thermoluminescent signals. These elements are used in different compounds as thermoluminescent materials (for example Al2O3, BaF2, BaSO4, CaF2, CaSO4, MgB4O7, MgO, Mg2SiO4 and MgF2). The presence of those elements in the surface contaminant can produce therefore new signals which interfere with the dosimetric signal of interest. These signals can be produced by means of excitation or processes like photo-, tribo-, chemiluminescence, etc. Additional investigations were carried out in order to identify the chemical form of those elements in the powder, and to recognise the way in which the interference occurs. It was found that the powder contains Al2O3 and CaF2, both well known for their thermoluminescent characteristics. Therefore, the reading of dosimeters contaminated with such a dust produces irregular dosimetric signals, the TL signals of Al2O3 and CaF2 and the responses of other thermoluminescent contaminants contained on the dosimeters according to environmental conditions of the client.
4. Simplified description of the method and its application A simple algorithm, which permits the identification of irregular glow curves in routine dosimetry, was developed. The basis of the algorithm is the differences in the shape of a regular and an irregular glow curve. Presently, the algorithm is being tested for its effectiveness in routine application. Different investigations were made, and an apparent problem is the existence of a chemical contaminant in the surface of some dosimeters. Now, the reasons for the irregularity have been studied and these spurious effects can be mainly eliminated by cleaning the dosimeters with methanol. Additionally, studies with the irregular curves recognised by application of the algorithm facilitate the identification of other unknown spurious effects and will thus help to eliminate interferences from chemical contaminations.
References [1] McLinlay AF. Thermoluminescence dosimetry. Medical Physics Handbooks, vol. 5. Bristol: Adam Hilger; 1981. p. 32 – 37. [2] Hoots S, Landrum V. Glow-curve analysis for verification of dose in LiF chips. Health-Physics-USA 1982;43(6):905 – 12. [3] Lawson BJ, Moleski M, Karandy R, Phillips CR, Champagne R, Winslow RJ. Harshaw/Bicron 8807 Environmental Dosimeter—Implementation Issues, TLD Users Symposium: Changes and Opportunities in Dosimetry. Las Vegas, 13 – 17 March 1995. Reprints of Presentations; 1995. [4] German V, Weinstein M, Ben-Shachar B, Pellet O. The contribution of Teflon to the thermoluminescent signal, Israel Radiation Protection Association (Israel). The Second Regional Mediterranean Congress on Radiation Protection, 20th Regional Congress of the Israel Radiation Protection Association, Tel-Aviv, 16 – 20 Nov. 1997. Program and Extended Abstracts 321 1997;39 – 42. INIS: 29-045616. [5] Weinstein M, Ben-Shachar B, German V. Determination of the contribution of thermoluminescent signals caused by disturbing effects, Israel Radiation Protection Association (Israel). The Second Regional Mediterranean Congress on Radiation Protection, 20th Regional Congress of the Israel Radiation Protection Association, Tel-Aviv, 16 – 20 Nov. 1997. Program and Extended Abstracts 321 1997;221 – 4. INIS: 29-045666.
122
V. Osorio Piniella et al. / J. Biochem. Biophys. Methods 53 (2002) 117–122
[6] Uray I. A background treatment for TLDs. Radiat Prot Dosim 1992;40(4):275 – 8. [7] German V, Weinstein M, Ben-Shachar B. Thermoluminescent signals caused by disturbing effects. Radiat Prot Dosim 1999;84(1 – 4):167 – 70. [8] Osorio Piniella V, Stadtmann H, Lankmayr E. A new algorithm to identify abnormal glow curves in TL personnel dosimetry. Radiat Prot Dosim 2001;96(1 – 3):139 – 41. [9] Duftschmid KE, Stadtmann H, Strachotinsky Ch, Winker N. A new type of spurious reading in TL personnel monitoring. Radiat Prot Dosim 1996;66(1 – 4):49 – 52.