Characterization of a water-based liquid scintillator for use in megavoltage radiotherapy beams

The measurement of the dose delivered in radiotherapy treatments is carried out using dosimeters that are often expensive to produce and sometimes toxic to humans and the environment, which leads to more complex and rigorous clinical manipulations. It is in this context that it is necessary to provide new types of scintillators that would no longer have these problems while having properties equivalent to those of human tissues. Thus, the following study presents the performance of a water-based liquid scintillator used at radiotherapy energies. The characteristics studied include the proportionality of the scintillation signal to the dose, the scintillation efficiency at two different energies as well as the identification of the Cherenkov portion of the signal for photon beams of 180 kVp, 6 MV as well as 18 MV. Spectral measurements of the scintillation solution and a solution of distilled water were acquired in order to isolate the contribution of the scintillation signal from the spectrum obtained, and then compared to a commercial scintillator, Ultima Gold. The signal exhibits a linear dose relationship with a correlation coefficient of 0.999 and lower scintillation efficiency than Ultima Gold.


Introduction
The need to measure accurately the dose for radiation treatments is one of the reasons why there is various kinds of detectors available in clinical settings.However, in the photon beam energy range where Compton scattering is dominant, most of these detectors do not have interaction properties that are similar to those of human tissues [1].Organic liquid compounds that are composed of a solvent and fluorophores possess this key advantage but often have a solvent that can be toxic and therefore require tedious manipulations [1].Water-based liquid scintillators (WbLS) have tremendous potential since the solvent is mainly composed of water and exhibit similar light emission characteristics as a standard liquid scintillator.In this work, we focus on the characterizing of a WbLS composed of 80% of water with less than 20% of the organic solvent linear alkylbenzene (LAB) and 2,5-diphenyloxazole (PPO).More specifically, this work investigates the relationship between the dose and the scintillation response as well as the spectral response of the detector and its variation with the energy of the incident beam.

Water based liquid scintillator description
The water based liquid scintillator (WbLS) was produced by the Neutrino and Nuclear Chemistry team at Brookhaven National Laboratory affiliated with the research team of Gabriel D. Orebi Gann.The liquid scintillator was given to our team for characterization.This dosimeter is composed of 80% of distilled water and 20% or less of an organic solvent named linear alkylbenzene (LAB).The fluorophores used are 2,5-diphenyloxazole (PPO) and represent less than 0.1% of the total composition of the liquid scintillator [3].

Irradiation of the scintillator and signal collection
All MV irradiations were performed with a medical linear accelerator TrueBeam (Varian, Palo Alto, USA).The scintillation signal was collected and guided through a clear optical fiber as seen in Figure 1.The spectra obtained from each irradiation was measured from the spectrometer QE-65 Pro (Ocean Optics, Dunedin, USA) with a fixed integration time of 25 s.It is to be noted that the spectra were not corrected for the wavelength dependencies of the fiber and the spectrometer.A field of 10x10 cm 2 was employed and the vial containing the solution to be analysed was placed at the linac isocenter.Water slabs were placed around the vial to act as a backscattering and build up medium and the vial was positioned at the depth of maximum dose for each energy used.For investigations using kV energies, the orthovoltage machine was used (Xstrahl, Camberley, United Kingdom).

Linearity of the scintillation signal
The WbLS signal linearity was investigated for a photon beam energy of 6 MV.The dose rate was kept constant, only the number of MU changed to vary the dose at the isocenter.There were no corrections applied for the production of Cherenkov light in this section since it did not affect the linearity of the signal.

Energy dependence of the signal
Three distinct irradiations at 6 MV, 10 MV and 18 MV were performed with the vial of WbLS.The dose was kept constant for all of the irradiations.To correct for the Cherenkov emission in MV energy range, a vial containing water was exposed to the same beams under the same reference conditions.In water, only Cherenkov emission is collected.The minimum energy required to produce this parasite signal depends on the refractive index of the irradiated medium.In water, this energy is 264 keV [1].Therefore, to obtain the signal without Cherenkov, irradiations at 180 kVp were performed and the signal without Cherenkov at MV energies was found using the hyperspectral method [2].The liquid scintillator Ultima Gold (Perkin Elmer, Waltham, Massachusetts, United States) was irradiated in the same conditions as for the WbLS for comparison.A blank irradiation using an empty vial was performed for each energy to subtract the possible Cherenkov emission produced in the optical light collecting fiber.

Linearity of the scintillation signal
The scintillation signal obtained for the WbLS is shown in Figure 2 as a function of the dose delivered.The signal obtained follows a linear trend with doses ranging from 0 to 300 cGy.The correlation coefficient R 2 obtained is 0.999.The dose was determined knowing the number of MU programmed to the vial that was irradiated at reference conditions.

Energy dependence of the scintillation signal
The spectral response of the detector is represented on Figure 3.The signal is decomposed into its two components, the Cherenkov and the scintillation portion.There is a non negligeable production of Cherenkov scintillation in the signal that needs to be canceled for the analysis of the pure scintillation signal.The scintillation signal of both the scintillators is shown in Table 1 as a function of the energy of the photon beam.Each value is normalized to the scintillation signal of the Ultima Gold at 6 MV.It is possible to see that the WbLS has a lower output than the liquid scintillator Ultima Gold at both energies.Ultima Gold has the highest energy dependance with an increased output of 5% at 18 MV compared to the signal obtained at 6 MV.WbLS presents a lower dependance with an output at 18 MV that is 0.6% more than the signal at 6 MV.

Conclusion
This study illustrates the advancements in the characterization of a new type of water-based liquid scintillator.While the WbLS produced a lower emission yield than the reference commercial LS, the signal is sufficient for dosimetry purposes.The energy dependence needs to be explored further, including lower photon energies as well as clinical electron beams.It should be noted that the concentration could be modified in order to observe the effect of the %LS loading on the signal obtained.With an optimal concentration and known characterization, this scintillator could be a potential candidate for the development of a new dosimeter that is less toxic and more clinically safe than those existing on the market.

Figure 1 .
Figure 1.Experimental set-up for the scintillation collection at MV energies.

Figure 2 .
Figure 2. Linearity of the scintillation signal for the water-based liquid scintillator as a function of dose for 6 MV irradiation.Error bars are smaller than the symbols presented.

Figure 3 .
Figure 3. Spectral measurements of scintillation that illustrates the Cherenkov component created in the water-based liquid scintillator at 6 MV compared to the scintillation signal without the presence of Cherenkov at 180 kVp.

Table 1 .
Signal ratio of water-based liquid scintillator and Ultima Gold for beam energies of 6 MV, 10 MV and 18 MV normalized to the scintillation output of the Ultima Gold at 6 MV.The Cherenkov contribution was removed in all cases.