Ultraviolet System - Experimental Setup - 基於光刺激發光技術用於Al2O3:C材料之劑量消光研究

Part 2- Experimental Setup

2.2.5 Ultraviolet System

Fig. 2-17. Experimental setups for visible light bleach.

For another qualitative bleach test, 5 parallel visible light sources were used (18W*5

=90W PHILIPS MASTER TL-D Super 80 /840 SLV) in the Fig. 2-17. The spectrum of each light source is centered at 420nm (blue), 560nm (Green) and 640nm (Red), combine 3-band fluorescence as shown in the Fig. 2-18. Under 18 cm of the light source, irradiation platform was designed for OSL badge bleach with UV filter to avoid UV effect (less than 40% below 370nm) shown in the Fig. 2-19.

Fig. 2-18. The emission spectrum of visible light source. (provided by the website of http://www.philips.com.tw)

UV filter

Part 2- Experimental Setup

Fig. 2-19.The transmission properties of UV filter [25].

3.1 Standard deviation

Chapter 3 Experimental Result

3.1 Standard deviation

Four standard badges (1mGy, 1Gy, 5Gy, and 10Gy) were used to evaluate the stability and reliability of Luxel^TM badge under 0.01mGy error reliability of Landauer POSL reader.

Usually, the one year background accumulation dose is approximately approached to 1 mGy and to map the value to 150 Convertible Value (Conv.

Val.) in the Landauer POSL reader. According to Fig. 3-3 and Fig. 3-4, badge has nearly 4.6%

error dose. It is acceptable due to the value is less than 0.05mGy which is half-month background accumulation dose. About 1Gy (=56000 Conv. Val.) badge, has nearly 5.3%

error dose. 5Gy (=316000 Conv. Val.) badge, has nearly 5.8% uncertainty. 10 Gy (=716000 Conv. Val.) badge has nearly 3%

fluctuation.

Luxel^TMbadges be exposed to standard dose from 1Gy to 10Gy and background dose

Using Landauer OSL reader continuously read out OSL intensity variation for 40 times

Draw the dose dependence of OSL intensity picture and calculate the standard deviation as the error bar of follow chart

Luxel^TMbadges be exposed to standard dose from 1Gy to 10Gy and background dose

Using Landauer OSL reader continuously read out OSL intensity variation for 40 times

Draw the dose dependence of OSL intensity picture and calculate the standard deviation as the error bar of follow chart

Fig. 3-1. Flow chart for Landauer OSL reader reliability test.

3.1 Standard deviation

Fig. 3-2. Average distribution of 1, 5, 10, 50, 100Gy for 40 Times OSL Read out.

Fig. 3-3. Standard deviation of four elements in background and 1Gy Badge.

Fig. 3-4. Standard deviation of four elements in 5Gy and 10Gy Badge.

3.1 Standard deviation

Standard Deviation of nonlinear dose dependence-50 Gy

The OSL signal (Conv. Val.) start to non-steady severe fluctuated when the dose was increased to extremely high dosage, i.e. above 50Gy. This means standard deviation have become worse. Fig. 3-5 is three 50Gy badges OSL intensity varies with 40 times read out. Apparently, two badges (XA00061705I & XA00838950A), in Fig. 3-5 have two layers to separate element 1 and element 2, 3, and 4. The main reason of this phenomenon could attribute to Luxel^TM badges shielding (chapter 2). The 50Gy (=436000 Conv. Val.) badge, has nearly 16% uncertainty so that the factor is significant parameter to distinguish 5~10Gy and above 10Gy to 50Gy. Although, the OSL intensity (Conv. Val.) of 5~10Gy and 10~50Gy is similar to 450000~550000 Conv. Val. comparing Fig. 3-4 and Fig. 3-5, the uncertainty is 5% and 16% respectively. For reliable data analysis, element-1 in XA00838950A is taken into consideration for numerical analysis reference point.

Fig. 3-5. 40 times readout distribution and standard deviation of four elements in three 50Gy Badges.

3.1 Standard deviation

Fig. 3-6. 40 times readout distribution of 5Gy & 10Gy.

Standard Deviation of nonlinear dose dependence-100 Gy

When the amount of dose increase to 100Gy, the OSL intensity (Conv. Val.) fluctuate more violently. In Fig. 3-7, the uncertainty of 100Gy badge is roughly 12.5%, 24%, 26% for XA00867979L, XA00254080V, XA008685817 respectively. Although the 100Gy OSL intensity seems very vague, the intensity (Conv. Val.) still locate in the same range of 5~10Gy. Therefore, there is a question left behind： How can we distinguish the badge is 5~10Gy or 50~100Gy by using the same Convertible Value?

Fig. 3-7. 40 times readout distribution and standard deviation of four elements in three 100Gy Badges.

3.2 Visible light effect

A visible light bleach system was designed (In chapter 2) to separate 5~10Gy and 10~50Gy badge. Each standard-dose badge was exposed under visible light tubes with UV filter for 3~10 min respectively. The result in Fig. 3-8 shows different decay speed during 10 min exposure. For example, XA008389609 (5Gy) & XA00061705I (50Gy) are at the beginning. For 10 min visible light bleach, the slope of 5Gy and 10Gy OSL variation is smaller than that of 50Gy and 100Gy. To further verify the validation of the study, exceed 10 min bleaching is needed for avoiding signal jumping, which is shown in Fig. 3-8. According to this bleach result, a critical slope value could be set to determine the dose in both high and low range. For easy slope recognition, exposure in 10 to 20 minutes is suggested.

Fig. 3-8. Visible Light Bleach effect for 5~100Gy Badge

The inset picture of Fig. 3-8 shows that the decay speed (slope) of 100Gy badges is slower than 5~10Gy. No matter how wide the range of OSL initial intensity it is, visible light bleach can still do the function to determine the badge 5~10Gy or above 10Gy.

0 100000 200000 300000 400000 500000 600000 700000 800000 900000

0 10 20 30 40

Time (min)

Element-1 Conv. Val. XA00254080V (100Gy)

XA00867979L (100Gy) XA008685817 (100Gy)

3.2 Visible light effect

Element-2 Conv. Val. (104 count)

Dose (Gy)

Element-1 Conv. Val. (104 count)

Dose (Gy)

Element-3 Conv. Val. (104 count)

Dose (Gy)

Element-4 Conv. Val. (104 count)

Dose (Gy)

Element-2 Conv. Val. (104 count)

Dose (Gy)

Element-1 Conv. Val. (104 count)

Dose (Gy)

Element-3 Conv. Val. (104 count)

Dose (Gy)

Element-4 Conv. Val. (104 count)

Dose (Gy)

Time = 0 min Time = 10 min Time = 20 min

Fig. 3-9. The Visible light Bleach effect for Element 1,2,3 and 4.

Figure 3-9 demonstrates that bleach effect in four elements for each badge is similar. There is no need to see four elements bleach effect in the following text. The first element will be use as experimental reference data for simulation analysis.

3.3 UV light effect

A UV light bleach result is displayed in Figure 3-10. Two background dose (nearly under 200 Conv. Val. = under 1mGy) badges were exposed to UV. The “step” in horizontal axis stands for 3 min UV exposure. Form “step1” to “step8”, a pump effect was discovered. “step8” to “step9” is that badges were laid on the table without UV exposure. Form “step9” to “step11”, another pump effect was discovered too. It means UV light for Luxel^TM Al2O3:C badge not only has “bleach effect” (Fig. 3-11), but also has

Fig. 3-10. UV light pumping effect for background dose badge. Note that each “step”

stands for 3 min UV exposure and “step8” to “step9” is 12 hours.

0 2 4 6 8 10 Before 1 Hour UV Bleach After 1 Hour UV Bleach

Fig. 3-11. UV light bleach effect for 1~10 Gy badge.

4.1 Set differential equation

Chapter 4 Simulation and Discussion

4.1 Set differential equation

Chapter 2 has already introduced how to use band diagram to describe electron behavior in the Al2O3:C and how to write down the ordinary differential equation (ODE) according to band diagram. As ODE is a dynamic system, choosing a proper numerical method must be considered. After that, tune the parameter in each ODE because each parameter has its own physical probability meaning. Finally, compare the simulation data with experimental data.

Checking if they match or not, if not we should modify band diagram or tune a new set of parameters, as shown in Fig. 4-1.

Put ODE into Matlab 2007 software and use “ode23” to solve this dynamic simulation Writing down proper

Ordinary Differential Equation (ODE)

Tune the parameter in the ODE and compare with experimental result

Put ODE into Matlab 2007 software and use “ode23” to solve this dynamic simulation Writing down proper

Ordinary Differential Equation (ODE)

Tune the parameter in the ODE and compare with experimental result

Fig. 4-1. The flow chart of simulation process.

4.1 Set differential equation

There are four steps (Fig. 4-2) or four strategies to describe electron behavior in the Al2O3:C. First, simulate Al2O3:C under high energy radiation explosion such as X-rays or beta-rays. Follow that, the system has to obey thermal equilibrium process. Electrons in this step should be distributed to another state by thermal energy. Theoretically, ODE dynamic system should be added with some thermal excitation term. Third, writing down a set of bleach ODE by adding bleach factors to simulate electron for pumping away electrons from the original states. Finally, same as bleaching step, POSL also need to pump away electrons in dosimetric trap. Thus, change bleach factors to POSL bleach factors. Then read out the concentration variance to calculation OSL intensity (so called “glow curve”).

Valence Band

Fig. 4-3. Schematic R. Chen’s multi-trap, multi-recombination-center energy model.

The used model is shown in Fig. 4-3, N1 which is an “active” dosimetric trap, meaning that the stimulating light is capable of releasing electrons from it, and N2, a competitor, which the electrons can be trapped, but the stimulating light cannot release

1. Material be bombarded with high energy 1. Material be bombarded

with high energy

Fig. 4-2. The steps of ordinary differential equation in Matlab 2007 software.

4.1 Set differential equation

electrons from it. Two recombination centers are assumed to exist, M1, radioactive center, and M2 a non-radioactive competitor. The set of six coupled differential equations are governing each process. The repeated simulation of R. Chen’s is in the figure 4-4.

Bombard with high energy radiation

Physical description Unit

M1 The concentration of radioactive hole centers (cm⁻³) M2 The concentration of non-radioactive hole centers (cm⁻³)

m₁ Instantaneous occupancy (cm⁻³)

m₂ Instantaneous occupancy (cm⁻³)

N₁ The concentration of the electron active trap state (dosimetric trap, shallow trap)

(cm⁻³)

N2 The concentration of the trapping state (cm⁻³)

n1 Instantaneous occupancy (cm⁻³)

n₂ Instantaneous occupancy (cm⁻³)

n_c The concentrations of the electrons in the conduction bands (cm⁻³) nv The concentrations of holes in the valence bands (cm⁻³)

4.1 Set differential equation

X The rate of production of electron-hole pairs, which is proportional to the excitation dose rate

(cm⁻³)(s⁻¹)

B1, B2

The trapping probability coefficients of free holes in Recombination centers 1 and 2

(cm⁻³)(s⁻¹)

Am1, A_m2

Recombination probability coefficients for free electrons with holes in centers 1 and 2,

(cm⁻³)(s⁻¹)

An1 The re-trapping probability coefficient of free electrons into the active trapping state N1 (dosimetric trap)

(cm⁻³)(s⁻¹)

An2 The trapping probability coefficient of the free electrons into the competing trapping state N2

(cm⁻³)(s⁻¹)

(X*T) If we denote the time of excitation by (T) and the rate of production of electron-hole pairs per cm³ by (X), then represents the total concentration of electrons and holes produced, which is proportional to the total dose imparted.

Gy, Sv

f A magnitude proportional to the stimulating light source

intensity. (s⁻¹)

Take out rate of production of electron-hole pairs X, comparing (3-5) with (3-11)

4.1 Set differential equation

Radiative Center Occupancy / Integral OSL

m10 m1f m10-m1f

Fig. 4-4. Repeat R. Chen’s simulation result. The right side of the figure is the simulation of integral OSL versus dosimetry (log scale) and is fully matched with the result of R Chen’s proposed (left side) [10].

4.2 Modification

4.2.1 Bombard with high energy radiation

The main OSL emission band at ~420nm (Akselrod et al., 1990 [26]; Markey et.,1995 [27]) is attributed to radioactive relaxation of excited F-centers, which are created after electron recombination with F+-centers (Springis et al., 1995 [28]). When F-center accepts 206nm UV light, it will convert back to F+ center (F-center + 206 nm Æ F+-center + e-). F+ center is the luminescence center in Al2O3:C [29]. It is one of dominate factors for OSL intensity. Thus back to R. Chen’s model, we modified Eq.3-3 &

Eq.3-4 to Eq.3-25 & Eq.3-26 respectively and added C m n and ₁ ₂ _c C m n into the ₂ ₂ _c equations. These two terms represent the probability that F-center (M2) would accept enough high energy and become F+-center (F-center + high energy (beta-rays) Æ F+-center + e-). The phenomena depends on m2 concentration (F-center) and electron concentration nc. Conversion rate C1 was set equal to C2 with opposite sign. This setting represents the 100% F-center conversion will change to F+-center. Figure 4-5 is C1 & C2 parameter simulation. This figure shows the F-center conversion must occur in the Al2O3:C. Comparing with Figure 4-5(b), experimental result, we chose C1 = C2 = 50 for the following simulation.

1 1 1

(

1 1

)

1 2

m c v c

d m A m n B M m n C m n

d t = − + − +

^(3-25)

2 2 2

(

2 2

)

2 2

m c v c

dm A m n B M m n C m n

dt = − + − −

^(3-26)

4.2 Modification

Relatice OSL Intesity / OSL Count

Dose (Gy) C1 = C2 = 100 C1 = C2 = 50 C1 = C2 = 25 C1 = C2 = 0

Fig. 4-5. (a)F-center conversion simulation result and (b) experimental data.

4.2.2 Optical Bleach

The mathematical term(−V n_f) ₁ was set into n1 state, same as in the set of equations for OSL process. This is base on the assumption that visible light only cleans electrons in the dosimetric trap and there is no enough energy to bleach electrons in deep trap. The simulation result is shown in Fig 4-6. Point “A “to “H” is the experimental data and they are roughly match our simulation result in Fig. 4-6, Fig. 4-7 and Fig. 4-8.

Fig. 4-6. Optical bleach simulation result with experimental data.

4.2 Modification

0 100000 200000 300000 400000 500000 600000 700000 800000

0 2 4 6 8 10

Dose (Gy) Element-2 Conv. Val. ^{Time = 0}

Time ~ 10 min Time ~ 20 min

Fig. 4-8. Detail bleach effect of optical bleach effect in 1~ 10Gy with simulation result.

Fig. 4-7. Simulation matches the OSL decay speed with different dose.

0 5 10 15 20

0 0.5 1 1.5 2 2.5

3x 10⁴

Time (min)

OSL Relative Intensity / OSL Count 5Gy

10Gy 50Gy 100Gy

4.2 Modification

4.2.3 Maximum Bleach

Fig. 4-9. Bleach effect with UV filter in the visible light system.

In the personnel dosimetry and environmental dosimetry, maximum bleach for badge is very important. Both of dosimetry estimate dose range in mGy range. Figure 4-9 and figure 4-10 are under visible light system with UV filter and without UV filter respectively. With UV-filter and 10 min exposure case, the dose of badge will decrease to 0.1 mGy (=10 Conv. Val.) as shown in Fig. 4-10. On the other hand, in Fig. 4-11 without UV-filter, the dose of badge will left 0.2 mGy (=20 Conv. Val.). This phenomenon describes UV light in the visible light tube will increase dose in the badge. In physical explanation, UV light source can not only bleach electrons trap center but also add extra electron into trap center. Therefore, ODE dynamic system should be modified as following:

1 1 1 1 1

( _f ) ( _v) _n ( ) _c

dn V n U n A N n n

dt = − + − − (3-28)

1 2 2 2

(

)

( )

dn U n A N n n

dt = − + −

(3-29)

1 1

( )(

1 1

)

m c v

dm A m n U M m

dt = − + −

(3-23)

4.2 Modification

For UV light could pump electrons from dosimetric trap and deep trap, (−U n_v) ¹was written into N1 state and N2 state in the ODE, where Uv is the coefficient of UV impact factor. On the other hand, UV light also could add electron into dosimetric trap and deep trap. Therefore, (U_v)(M₁−m₁) was put into M1 state. In order to determine the simulation parameter Uv and Vf, , we calculate minimum photon unit by bleach light source spectrum intensity (Fig. 2-16&2-18) and then estimate photon number of wavelengths. We assume 400nm is a critical wavelength between bleach and pump effect. Moreover, we set 50 to 50 percentages for UV pump and bleach effect respectively. Finally, set the simulation parameter of bleach photon number of visible light as 0.5 and others proportional to this value. The simulation result is shown in Fig.

4-12.

Table 4-1. Calculation of lowest common multiple for bleaching photon 420nm = 2.95ev, 560nm=2.21ev, 640nm=1.93ev, 1ev = 1.6*10¹⁹J 90W*60sec=5400J

Visible

Lamp 5400J/(80*3.1+200*2.95+400*2.21+350*1.93)/10¹⁹=2.25714*10¹⁹ (Minimum photon unit for visible light)

350nm=3.54ev, 1ev = 1.6*10¹⁹J 4w*60sec=240J

UV Lamp

240J/3.54/10¹⁹=6.7796*10²⁰

(Minimum photon unit for UV lamp)

154

Fig. 4-10. Visible Light bleach without UV filters. (Each step has 10 min time gap)

4.2 Modification

Table 4-2. Simulation parameter dependence Photon

Number UV (<400nm) Visible(>400nm)

UV pump (50%) UV bleach (50%) 40*2.25*10¹⁹=9.02*10²⁰ 40*2.25*10¹⁹=9.02*10²⁰

(200+400+350)*

2.25*10¹⁹

=2.14*10²²

Simulation Simulation Simulation

Visible Lamp

0.021 0.021 0.5

0.5*6.77*10²⁰=3.38*10²⁰ 0.5*6.77*10²⁰=3.38*10²⁰

Simulation Simulation

Relative OSL Intensity / OSL Count

Time (min)

Normalized Relative OSL Intensity / OSL Count

Time (min)

Uv = 0.021, Vf = 0.5+0.021,Without UV filter Uv = 0.004, Vf = 0.504,With UV filter

Fig. 4-11. Bleach effect will achieve to a saturation point.

Fig. 4-12. Different light source has different saturation point.

The simulation results are shown in Figure 4-11 and Fig. 4-12. It successfully explains UV bleach effect with pumping function and the OSL intensity saturate at the same place. In Fig. 4-11, different n1 initial concentration represents different

experimental condition in Fig. 4-9. The other simulation result in Fig. 4-12 demonstrates different bleach light source has different saturation effect. This simulation explain even though Visible light tube is for bleach usage, but it also has pumping factors to achieve low saturation. Therefore, UV filter usage in visible light system is necessary for deep bleach in OSL badge.

4.2 Modification

4.2.4 Predictions by this model

What if the stimulated light source changes from Green light to UV light?

There suppose to have possibilities that UV light could pump deep trap electron n2. Therefore, using UV light as a stimulated light source not only pump shallow trap electrons (dosimetric trap n1), but also pump deep trap n2 as well, as shown in Fig. 4-13.

For UV light source has different impact factors for dosimetric trap and deep trap, an impact factor “d” was set to deal with this question. Impact factor “d” was set from zero to one. When “d” is zero or one that means stimulated light source can only pump electrons in dosimetric trap or deep trap respectively. Therefore, OSL intensity only attribute to one trap state. When “d” is equal to “0.5”, in the middle of zero and one, which means stimulated light source can pump electrons in both traps and it is a pumping competition. Therefore, stimulating light source impact factor f1 and f2 (in the Figure 3-24) are set (-f)(1-d) and (-f)(d), in the Eq. 3-13.&Eq. 3-14. “-f” is correlated with stimulating light source and “d” is associated with one trap stimulated or two trap stimulated.

The simulation results are shown in Fig. 4-14. When “d=0%”, the result is again back to R. Chen’s paper (or Fig. 4-4). When “d=50%”, two trap stimulated competition occur, the OSL intensity is still similar to one trap stimulation. This is an important prediction for POSL using UV light source. Because POSL only consume little amount of

Valence Band

Fig. 4-13. Two trap stimulated competition band diagram [30].

4.2 Modification

electrons in the traps and two traps concentration will not run out completely, thus pumping probabilities will follow with two trap concentration ratio condition. In 2009, We proposed a quantum selection rule for OSL and indicated pumping probabilities depend on two traps energy level ratio with concentration ratio [30]. For long time stimulation, the pumping probability of two traps should reduce to 1:1 which is the same function in this simulation “d=50%”. In the extreme case “d=100%”, shallow trap ran out of electron concentration and only deep trap can do the function. In reality, this phenomenon can only take place in the pre-bleach process. Dosimetric trap (shallow trap) had been already bleached out by pre-heat or optical bleach before UV-Pulse OSL measurement.

According to this mathematical model prediction, pre-bleach shallows process seems to enlarge linear dose dependence from 10Gy to 50 Gy.

0 20 40 60 80 100

-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

Relative OSL Intensity (104 count)

Dose (Gy)

d = 0%

d = 50%

d = 80%

d = 90%

d = 100%

Fig. 4-14. OSL intensity for two traps stimulates competition.

Chapter 5 Conclusions

Although maximum precision measurement of Landauer POSL Reader with Luxel Al2O3: C badge is 10Gy, the high dose exceeds 10Gy measurement still can finish roughly by the decay slope of optical bleach process and standard deviation calculation in our experiment (Figure 5-1). We also find out the modification model to explain this bleach decay phenomena successfully. Those two methods enlarge the estimate scale for the Landauer reader with Luxel Al2O3: C badge in the case of nuclear event or nuclear emergency.

Suspected Badge

Fig. 5-1 The proposed flowchart to distinguish the over dose and under safety margin situation.

Chapter 5 Conclusions

Using the UV lamp as bleach light source has not only bleach function, but also pump effect [20]. Therefore, we use commercial visible light source with UV filter instead of UV lamp to avoid pump effect [16]. Moreover, in our optical bleach simulation, the saturation point with UV filter and without UV filter is different. These simulations totally match experimental evidence. That is the saturation point for maximum bleach could be attributed to balance of the bleach effect and pump effect.

So that we could make a prediction that the higher energy photon incidents will cause the pumping effect and bleaching effect simultaneously. When the competition approaches the equilibrium, the higher residue of the electrons will be staying the trapped level so that we could make a fine tune to set the electrons quantitatively trapped in the desired energy level.

When F-center accepts 206nm UV light, it will convert back to F+ center (F-center + 206 nm Æ F+-center + e-) [20]. Therefore, we assumed that F-center converts to F+-center (F-center + high energy (beta-rays) Æ F+-center + e-) should occur in Luxel Al2O3: C badges. This assumption successfully describes the phenomena of OSL intensity saturation in high dose in our simulation.

By modified simulation result, changing light source energy from green LED to higher energy light UV lamp in POSL system seems no apparent signal change and maintains its original OSL curve. Therefore, using two trap model by changing light source energy seems provide another idea for new type of OSL reader. Moreover, according to simulation result in UV stimulation, clean shallow trap electron concentration left deep trap dose the function seems to enlarge linearity dose dependence from 10Gy to 60Gy in Al2O3: C. This phenomena provided us an idea in

在文檔中基於光刺激發光技術用於Al2O3:C材料之劑量消光研究 (頁 39-0)