Fricke-gel dosimeters proved to be suitable tools to perform 3D radiotherapy pre-treatment dosimetry. The tissue equivalent gel matrix helpsto preserve the spatial information of the dose. Several gel matrices proved to be suitable for dosimetric purposes. The influence of the gel matrices on both system dose response and diffusion processwasinvestigated. Three gel matrices were considered: Gelatinfrom porcine skin, Agaroseand polyvinyl alcohol (PVA) cross-linked with glutaraldehyde(GTA). In these systems, Xylenol-Orange (XO), an iron(III)chelator, forms red-colored complexeswith Fe3+ that eases the optical determination of the dose. However, the dose evaluationresults to be affectedbydifferent XO-Fe3+ complexes that absorb at different wavelengths. In particular,this phenomenon influences the dose response, the calibration curveand the dose threshold.Therefore, a deeper study of the XO-Fe3+ speciation mechanism could lead to a more accurate evaluation of the dose. A novel procedure, based on a laser-beam irradiation, was implemented for the diffusion process evaluation. The diffusion coefficients were calculated for the three gel matrices tested. PVAmatrix proved to highly limit the diffusion with respect to the other matrices. Further investigations are needed to verify the influence of XO-Fe3+complexeson the diffusion phenomenon.

1. International Journal of Engineering Science Invention
ISSN (Online): 2319 – 6734, ISSN (Print): 2319 – 6726
www.ijesi.org ||Volume 5 Issue 7|| July 2016 || PP. 48-54
www.ijesi.org 48 | Page
Gel matrix dependence on the dose response properties and
diffusion phenomena of Fricke-based gel dosimeters
Giulia Maria Liosi1
, Simone Lazzaroni2,3
, ElenaMacerata1
,Daniele Dondi2
1
Department of Energy, Nuclear Engineering Division, Politecnico di Milano, Milano, Italy
2
Department of Chemistry, University of Pavia, via Taramelli 12, 27100 Pavia, Italy
3
Istituto Nazionale di Ricerca Metrologica (INRIM), via Taramelli 12, 27100 Pavia, Italy
Abstract: Fricke-gel dosimeters proved to be suitable tools to perform 3D radiotherapy pre-treatment
dosimetry. The tissue equivalent gel matrix helpsto preserve the spatial information of the dose. Several gel
matrices proved to be suitable for dosimetric purposes. The influence of the gel matrices on both system dose
response and diffusion processwasinvestigated. Three gel matrices were considered: Gelatinfrom porcine skin,
Agaroseand polyvinyl alcohol (PVA) cross-linked with glutaraldehyde(GTA). In these systems, Xylenol-Orange
(XO), an iron(III)chelator, forms red-colored complexeswith Fe3+
that eases the optical determination of the
dose. However, the dose evaluationresults to be affectedbydifferent XO-Fe3+
complexes that absorb at different
wavelengths. In particular,this phenomenon influences the dose response, the calibration curveand the dose
threshold.Therefore, a deeper study of the XO-Fe3+
speciation mechanism could lead to a more accurate
evaluation of the dose. A novel procedure, based on a laser-beam irradiation, was implemented for the diffusion
process evaluation. The diffusion coefficients were calculated for the three gel matrices tested. PVAmatrix
proved to highly limit the diffusion with respect to the other matrices. Further investigations are needed to
verify the influence of XO-Fe3+
complexeson the diffusion phenomenon.
Keywords: Diffusion, Fricke-gel dosimeter, Gel matrix, Optical spectrometry, Radiation dosimetry
I. Introduction
Some open issues still need to be faced in order to optimize and standardize the Fricke-gel dosimeter so
as to obtain an effective tool for pre-treatment 3D dosimetry[1 - 5]. Such dosimeter is based on the dose
dependent oxidationof ferrous ions (Fe2+
) dispersed into a tissue equivalent gel matrix which preserve the spatial
information of the absorbed dose. Different gel matrices have been studied and proved to be suitable for
dosimetric purposes [2,4]. The ferric ions (Fe3+
) concentration is linearly related to the dose by means of a
proper calibration factor. The Fe3+
ions concentration can be measured by the difference in absorbance at
304 nm before and after irradiationThe dose response is therefore defined as the ratio between the difference in
absorbance and the total absorbed dose. [2]. A ligand, the Xylenol Orange (XO), proved to be suitable to chelate
the Fe3+
ions leading to the formation of complexes which mainly absorbs in the visiblerange, at 585 nm [6].In
this region,gel matrices do not present relevant absorption and thusthe accuracy and the precision of the dose
evaluation are enhanced. Moreover, this ligand reducesthe mobility of the Fe3+
so as to limit the diffusion and
better preserve the spatial information of the absorbed dose [7-13].
However, it is known that few open issues still need to be faced in order to i) optimize the system
properties in terms of sensitivity, stability and reproducibility, and ii) limit the diffusion phenomenon. In
addition, the system calibration procedure still need to be standardized. In fact, an apparent dose threshold is
reported in literature [14]. The dose threshold is a sort of non-zero crossing of the calibration line, whichaffects
the accurate evaluation of the absorbed dose, especially at low doses. Different linear calibration curves and
dose thresholds are obtained at different measurement wavelengths. Moreover, a different dose response has
been observed as a function of the total dose absorbed by the system [15].Some authors related the these
properties to the presence of different complexes between XO and ferric ions which possess a maximum of
absorption at different wavelengths: Fe3+
1-XO2at 480 nm; Fe3+
1-XO1 at 540 nm; Fe3+
2-XO1 at 590 nm [1,16].
Therefore, since by increasing the dose, i.e. the ferric ions concentration, the concentration of the different
complexes changes,different calibration curvesareobtainedat different wavelengths. In addition, effects on the
dose response, i.e.the ratio between the absorbance at 585 nm and the total absorbed dose,of samplessubjected
to multiple subsequent irradiationswith5 Gysteps have been observed at 585 nm[3,15].
Several works, reported in literature, studiedthe diffusion process for a single gel matrix (either Gelatin,
Agarose or PVA) as a function of the different chemical composition of the dosimeters [7 - 13]. However, these
results are scattered and affected by uncertainty due to the lack of a standardized experimental method. In fact,
different sample shape and preparation procedure, temperature condition, irradiation, measurement and data
analysis could lead to results not directly comparable.

2. Gel matrix dependence on the dose response properties and diffusion phenomena of Fricke-based ....
www.ijesi.org 49 | Page
In this work, a systematic analysis has been conducted to investigate the influence of the gel matrix on
both the system dose response and the diffusion process. In particular, Gelatin from porcine skin, Agarose and
PVA matrices have been studied. In addition, the effect of different gel matrices on the response to subsequent
irradiations of 5 Gy has been tested. Finally, a novel procedure based on the laserirradiation has been developed
to quickly and reliably test the diffusion phenomenon for the three different gel matrices.
II. Material And Methods
The Fricke-gel dosimeters consist of an acid aqueous solution of Fe2+
ions and XO infused in a
gelatinous matrix. In particular, the chemical composition of the aqueous solution is: 0.5mM Ferrous
Ammonium Sulphate (FAS, Sigma Aldrich F3754), 25 mMsulphuric acid (Carlo Erba Reagent grade 96%
pure), 0.165 mMXylenol Orange (Sigma Aldrich 33825). The gel matrices studied are: 3% w/vGelatin from
porcine skin (300 bloom gel strength, Sigma Aldrich G2500); 1%w/v Agarose (Sigma Aldrich A0576); 10%
w/v PVA-GTA (Poly(vinyl alcohol) Sigma Aldrich 341584 cross-linked with 6.62 mM of Glutaraldehyde
solution 25%Sigma Aldrich G6257). Fricke-gel dosimeters were prepared according to thestandardreceipts
reported in literature[17 -19]. In addition, in order to study the absorption properties of the gel matricesseveral
gelsamples, without the incorporation of the Fricke solution, were prepared. The final solution, till in liquid
phase,was poured into spectrophotometriccuvette (1.2x1.2x4.5 cm3
). Samples were stored in the dark at 7°C,
after a storage of 6 hours at room temperature.
The sample irradiations were performed at least 24 hours after the end of the gel preparation to ensure a
complete gelation, with a 137
Cs irradiator at the Medical Physics Unit (Fondazione IRCCS Istituto Nazionaledei
Tumori di Milano, Italy). The dose rate (0.113 Gy/sec) was measured by means of thermoluminescent
dosimeters (TLDs) placed in the same configuration of the gel samples.
All the absorption spectra were acquired with respect to a water reference sample by means of a
spectrophotometer (Perkin Elmer, Lambda 650) in the range 400-700 nm, 30 minutes after irradiation.
In order to characterize the system dose response, several samples, prepared with the three different gel
matrices, were irradiated in the range 0-40 Gy.Moreover, to investigate the effect of gel matrices on subsequent
irradiation, some Fricke-gel dosimeters were divided into two groups (A and B) irradiatedat step of 5 Gy using
different temporal sequences with time steps of 24 and 48 hours respectively.
In order to evaluate the diffusion, Fricke-gel layer dosimeters (9x9 cm, 2 mm thick) were prepared
according to a standardize procedure [19,20]. To calibrate the system, layers were irradiated at several doses in
the range 0-20 Gy. The analysis was performed by means of Nashuatec MPC2050 scanner 30 minutes after
irradiation. Color RGB images were acquired (600dpi). Therefore, the images weresplitted in three channels:
red, blue and green. By means of a proper standardized images manipulation [19, 20], it is possible to calculate
the Optical Density Difference (ODD). The ODD calculated for the green channel proved to be linearly related
with the dose.To evaluate the diffusion phenomenon, layers were irradiated with a laser beam (500 mW,
405 nm). During the irradiation, a photostationary equilibrium is obtained within minutes and the concentration
of complex resulted equivalent to 5 Gy. Filled circles with a diameters of 5 mm were drawn with the laser on the
layers.Layers were analyzed every 30 minutes, starting from 30 minutes after irradiation up to 2.5 hours. In
order to evaluate the diffusion coefficients, Gaussians were fitted to the measured profiles [21]. The diffusion
coefficients could be inferred through the relation (1) between the Gaussian and the Fick's law:
𝛿2
𝑡 = 2𝐷𝑡 (1)
where  is the standard deviation of the Gaussian fit, 𝐷 the diffusion coefficient and 𝑡 the time of analysis.
III. Results
3.1Dose response
The background absorbance of the three gel matrices considered has been compared as reported inFig
1. Agarose shows an higher absorption due to scattering phenomenon. In fact it is known that, while the Gelatin
from porcine skin has good optical properties, Agarose is translucent [22]. For this reason, all the Agarose
samples have been measured with respect to the gel matrix itself so as to properly subtract the background. The
absorption properties of both Gelatin and PVA matrices proved not to affect the spectrophotometer analysis in
the visible range of interest (400-700 nm).

3. Gel matrix dependence on the dose response properties and diffusion phenomena of Fricke-based ....
www.ijesi.org 50 | Page
Fig 1 Absorption spectra of the gel matrices without the inclusion of the Fricke solution. Absorption spectra
were acquired with a water sample as reference.
The spectra of non-irradiated samples made with the three different gel matrices are reported in Fig 2.
The initial signal is related to the spontaneous oxidation of ferrous ions occurring during the sample preparation.
The different chemical composition of the three matrices tested leads to different initial ferric ions concentration
due to a different oxidation rates depending on the matrix by itself and on the sample preparation. This
phenomenon reduces the dynamic range of the PVA sample to 0-25 Gy, while Gelatin and Agarose samples
shown a dose response linearly related with the dose up to at least 40 Gy.
Fig 2 Spectra of non-irradiated samples made with the three different gel matrices tested.
The spectra of the dose response, i.e. the ratio between the difference in absorbance (before and after
irradiation) and the total absorbed dose, are reported in Fig 3, Fig 4 and Fig 5 for PVA, Gelatin from porcine
skin and Agarose respectively.It is clearly visible a different behavior among the different gel matrices. In fact,
for both Gelatin and Agarose samples, the absorption spectra present a shape that varies with the dose and, in
particular, a different dose response as a function of the total absorbed dose. In fact, the absorbance at 585 nm
normalized on the total absorbed dose increases for higher doses. With PVA samples, the absorption spectrum
shape is instead not dosedependent and a lower dose response dependence (i.e. better linearity with the total
dose) is visible. This phenomenon can be attributed to the initial higher Fe3+
ions concentration of PVA samples
as shown in Fig 2. The higher ferric ions concentration leads to the formation of more Fe3+
2-XO1 which mainly
absorbs at 590 nm, increasing the dose response evaluated at 585 nm. In fact, while the concentration of Fe3+
increases, the equilibriums of different Fe3+
-XO species shifts toward the formation of different complexes:
Fe3+
1-XO2 + Fe3+
⇌ 2 Fe3+
1-XO1 (eq. 1)
Fe3+
1-XO1+ Fe3+
⇌ Fe3+
2-XO1 (eq. 2)
These phenomena affect the calibration curves obtained (Fig 6) because the three complexes absorb at
different wavelength. As a consequence, three different sensitivities and dose thresholds can be observed. As
previously described, the PVA present a lower dynamic range and, in fact, the samples irradiated with a dose
higher than 30 Gy present saturation. Moreover, due to the higher Fe3+
ions concentration and the consequent
higher Fe3+
2-XO1 complexes concentration, no dose threshold can be observed for PVA samples. When the

4. Gel matrix dependence on the dose response properties and diffusion phenomena of Fricke-based ....
www.ijesi.org 51 | Page
calibration curves for Gelatin and Agarose samples are evaluated after the subtraction of the absorption of the
sample irradiated at 15 Gy instead of the non-irradiated one, no dose threshold is observed (Fig 7). In fact,
higher is the dose, higher is the Fe3+
ions concentration and thus the Fe3+
2-XO1 concentration and consequently
higher is the contribution in the absorption spectrum at 585 nm. Contrarily to the previous calibration (Fig 6),
with this calibration curve it is possible to accurately measure low doses also with Gelatin and Agarose samples.
Consequently, in order to accurately measure very low doses without the dose threshold effect it is necessary: i)
to perform a pre-irradiation of the samples with a dose that must be identified for each preparation and chemical
receipt, or alternatively ii) to deeper study the speciation mechanism of XO and Fe3+
ions. In this way it would
be possible to develop a proper calibration procedure which not simply correlate the absorbance at one
wavelength to the dosebut, through a proper deconvolution of the absorption spectra, correlate the actual Fe3+
ions concentration to the dose.
Fig 3 Absorption spectra, resulting from the subtraction of the non-irradiated sample spectrum, normalized with
respect to the total absorbed dose for PVA samples irradiated in the range 10-40 Gy.
Fig 4 Absorption spectra,resulting from the subtraction of the non-irradiated sample spectrum, normalized with
respect to the total absorbed dose for Gelatin samples irradiated in the range 10-40 Gy.
Fig 5 Absorption spectra, resulting from the subtraction of the non-irradiated sample spectrum, normalized with
respect to the total absorbed dose for Agarose samples irradiated in the range 10-40 Gy.

5. Gel matrix dependence on the dose response properties and diffusion phenomena of Fricke-based ....
www.ijesi.org 52 | Page
Fig 6 Calibration curve Abs (585nm) vs Dose, obtained after the subtraction of the absorption of the
non-irradiated sample, for the three different gel matrices tested.Error bars are included in the marker size.
Fig 7 Calibration curve Abs (585nm) vs Dose, obtained after the subtraction of the absorption of the sample
irradiated at 15 Gy, for both Agarose and Gelatin gel matrices. Error bars are included in the marker size.
The effects of multiple subsequent irradiations have been evaluated. The results obtained are reported
in Fig 8 for both PVA (left) and Gelatin (right) samples. Similar results between Gelatin and Agarose samples
were obtained. The dose response is reported with respect to the number of subsequent 5 Gy irradiations.
Despite of the gel matrix, it can be observed that the dose response is dependent on the number of subsequent
irradiations, i.e. the total absorbed dose, but not on the irradiation time. In fact no difference can be observed
between groups A and B which were subjected to different temporal sequences of irradiation. It is consistent
with the previous data and is related to the different absorption wavelength of the Fe3+
-XO complexes. For PVA
samples instead, since the Fe3+
ions concentration is high even in low doses, the formation of Fe3+
2-XO1 is
favored, and thus the dose response results to be constant. On the opposite, coherently with the previous results,
the dose response of both Gelatin and Agarose sample increases with the number of subsequent irradiations, i.e.
with the total absorbed dose, and present a saturation trend. Therefore, considering the dose response at a fixed
wavelength could lead to incorrect conclusions on the system stability and response to fractionated doses due to
the presence of different Fe3+
-XO complexes.
Fig 8 Dose response as a function of the number of subsequent irradiationfor PVA (left) and Gelatin (right)
samples irradiated at step of 5 Gy with 2 different temporal sequences. For simplicity of reading, all the values
are reported normalized to the response at the first irradiation.

6. Gel matrix dependence on the dose response properties and diffusion phenomena of Fricke-based ....
www.ijesi.org 53 | Page
3.2Diffusion measurement
A novel procedure for the evaluation of the diffusion phenomenon, based on the laser irradiation of
Fricke-gel layers, has been implemented as follows. In order to validate the measurement procedure of the
Fricke-layers with the Nashuatec MPC2050 scanner, Fricke-gel layer-samples were irradiated in the range 5-20
Gy (calibration is reported in Fig 9).
Fig 9 Calibration curve ODD vs Dose obtained for Gelatin layers dosimeters irradiated in the range 5-20 Gy.
To evaluate the diffusion phenomenon, filled circleswith a diameter of 5 mm were drawn with a laser-
beam at an equivalent dose of 5 Gy. The experimental (left) and Gaussian fitted (right) profiles for Gelatin
samples are reported inFig 10. Similar results were obtained for Agarose and PVA samples, not reported for
sake of brevity. Thanks to the equation (1) the diffusion coefficients, reported in Table 1, were calculated. The
results obtained are coherent with literature [7 - 13]. The PVA matrix crosslinked with GTA proved to limit the
diffusion process more than Gelatin and Agarose matrices.
Fig 10 Experimental (left) and Gaussian fitted (right) profile obtained for Gelatin layer at different time interval
after irradiation. Similar results were obtained for PVA and Agarose samples, not here reported for sake of
brevity.
Table 1 Diffusion coefficient calculated for the three gel matrices tested
Gel matrix Diffusion coefficient (mm2
h-1
)
Gelatin from porcine skin 0.77±0.04
Agarose 0.57±0.03
PVA 0.38±0.03
IV. Conclusion
In this work the effect of the gel matrix on both the dose response and the diffusion phenomenon of the
Fricke-gel dosimeter have been investigated. The absorption properties of the system resulted to be affected by
artifactsrelated to thepresence ofdifferent XO-Fe3+
complexes, having different absorption spectra.
A novel procedure based on laser-beam irradiation was implemented in order to study the diffusion
phenomenon. The diffusion coefficients calculated are in accordance with the data reported in literature.
Moreover, for the reasons above mentioned, also the diffusion could be affected by the presence of different
complexes.

7. Gel matrix dependence on the dose response properties and diffusion phenomena of Fricke-based ....
www.ijesi.org 54 | Page
For this reason a deeper study of the speciation mechanism and a complete investigation of the whole
absorption spectra are necessary to avoid inaccurate measurements of both the total dose and the diffusion,
especially at low doses.
Acknowledgements
The authors wish to thanks the Medical Physics Unit at the Fondazione IRCCS Istituto Nazionaledei
Tumori di Milano for the irradiation campaign. This work has been supported by a co-funded Research Project
within the National Research Program of Relevant Interest (PRIN 2010) (Research Project no.
2010SNALEM_002).
References
[1]. Lepage M and Jordan K, 3D dosimetry fundamentals: gels and plastics, Journal of Physics: Conference Series 250, 2010,
doi:10.1088/1742-6596/250/1/012055
[2]. Schreiner L J, Review of Fricke gel dosimeters, Journal of Physics: Conference Series 3, 2004, 9-21, doi:10.1088/1742-
6596/3/1/003
[3]. Jordan K and Battista J, Dose response of ferrous-xylenol orange gels: the effects of gel substrate, gelation time and dose
fractionation, Journal of Physics: Conference Series 3, 2004, 232–235, doi:10.1088/1742-6596/3/1/035
[4]. Doran S J, The history and principles of chemical dosimetry for 3Dradiationfields: Gels, polymers and plastics, Applied
RadiationandIsotopes67, 2009,393–398, doi:10.1016/j.apradiso.2008.06.026
[5]. Baldock C, Historical overview of the development of gel dosimetry: apersonal perspective, Journal of Physics: Conference Series
56, 2006,doi:10.1088/1742-6596/56/1/002
[6]. Appleby A and Leghrouz A, Imaging of radiation dose by visible color development in ferrous-agarose-Xylenol Orange
gels,Medical Physics 18,1991, 309, doi: 10.1118/1.596676
[7]. Baldock C et al, Experimental determination of the diffusion coefficient in two-dimensions in ferrous sulphate gels using the finite
element method, Australasian Physics & Engineering Sciences in Medicine, 2001, 24:19, doi:10.1, 007/BF03178282
[8]. Olsson L E et al, Diffusion of ferric ions in agarosedosimeter gel, Phys. Med Biol. 37, 1992 2243–52
[9]. Rae W I D et al, Chelator effect onion diffusion in ferrous-sulphate doped gelatin gel dosimeters as analyzed by MRI, Med Phys,
23, 1996, 15–23
[10]. Balcolm B j et al, Diffusion in Fe(II/III)radiation dosimetry gels measured by MRI,Phys. Med. Biol. 40, 1995, 1665–76
[11]. Harris P Iet al, A method for determining the diffusion coefficient inFe(II/III) radiation dosimetry gels using finite elements Phys.
Med Biol. 41, 1996, 1745–53
[12]. Chu W C and Wang J, Exploring the concentration gradient dependency of the ferric iondiffusion effect in MRI-Fricke-infused gel
dosimetry, Phys. Med Biol. 45, 2001, L63–4
[13]. Penev K I and Mequanint K, Controlling sensitivity and stability of ferrous–xylenolorange–gelatin 3D gel dosimeters by doping
withphenanthroline-type ligands and glyoxal, Phys. Med. Biol. 58, 2013, 1823–1838
[14]. Babic S et al, An apparent threshold dose response in ferrous xylenol-orange gel dosimeters when scanned with a yellow light
source, Phys. Med. Biol. 53, 2008, 1637–1650, doi: 10.1088/0031-9155/53/6/009
[15]. Liosi G M et al, Study of Fricke-gel dosimeter calibration for attaining precise measurements of the absorbed dose, 4th
International Conference on Advancements in Nuclear Instrumentation Measurement Methods and their Applications (ANIMMA),
2015, IEEE, doi: 10.1109/ANIMMA.2015.7465581
[16]. Gay C et al, Determination of Iron in Solutions with the Ferric Xylenol Orange Complex, Analytical Biochemistry 273, 1999, 143–
148, doi: 10.1006/abio.1999.4207
[17]. Liosi G M et al.,Effects of the pre-irradiation storage procedure on the dose response of a Fricke xylenol orange gel dosimeter,
NUKLEONIKA 60(3), 2015, 603-607 , doi: 10.1515/nuka-2015-0099
[18]. Xiao R et al, Preparation and characterization of FPGX hydrogel dosimeters, Nuclear Science and Techniques 21, 2010, 60–62,
doi: 10.13538/j.1001-8042/nst.21.60-62
[19]. Gambarini G et al, Characterization of Fricke-gel Layers for AbsoluteDose Measurements in Radiotherapy, 2nd International
Conference on Advancements in Nuclear Instrumentation Measurement Methods and their Applications (ANIMMA),
IEEE,doi:10.1109/ANIMMA.2011.6172969
[20]. Gambarini G et al, Study of light transmittance from layers of Frickex-Xylenol-orange-gel dosimeters having different composition
and analyzed with various modalities, Nucl. Instrum.Method B 213, 2004, 321–324, doi:10.1016/S0168-583X(03)01636-7
[21]. De OlivieraL N et al, Determination of diffusion coefficient in Fricke Xylenol gel dosimeter after electronbeam
bombardment,Surface & Coatings Technology 203, 2009, 2367–2369, doi:10.1016/j.surfcoat.2009.03.044
[22]. Davies J B and Baldock C, Sensitivity and stability of the Fricke–Gelatin–Xylenolorange gel dosimeter, Radiation Physics and
Chemistry 77, 2008, 690–696, doi:10.1016/j.radphyschem.2008.01.007