by:?EOB?R[1?exp(??TEOB)] (2)
2.2 Neutron monitoring
TMThe MINItrace
1818Ftarget system is designed for production of 18Fvia the O(p,n)18Freaction. As a result, every successful interaction between a proton and the target nucleus results in the emission of a neutron. It gives a direct relationship that there will be the same amount of neutron produced as that of
TM18Fproduct during beam irradiation. Although the MINItracecyclotron is designed to be a self-shielded system, it is still
possible to monitor the neutron release from the target in real time outside the cyclotron by
33using a high-sensitivity moderated Heneutron detector. The Heproportional counter has
3a dimension of 2.54cm in diameter and 25.25cm in active length, filled with Hegas in a
pressure of 0.54Mpa. The thermal neutron sensitivity of the bare tube is about 32cps per unit neutron fluence rate. The detector was wrapped up by a 6.5cm thick cylindrical polyethylene to increase the response to fast neutrons. The counting rate of the neutron monitor is
1818proportional to the reaction rate of theO(p,n)Freaction, which the EOB activity can be
derived from. By observing the neutron counting rate during beam irradiation, we could monitor the process of target irradiation and provide instant information to operators.
3. Results and discussions
3.1Production efficiency
There are over 2100 batches of routine operation records accumulated in the log file of Shin Kong Memorial Hospital. The EOB activity was measured at the entrance of the synthesizer by a built-in solid state detector when the irradiated Owater had been transferred from the target site to the synthesizer. The EOS activity was measured by the Atomlab 300 dose calibrator at the end of synthesis. The ratio of the measured EOB activity to the calculated EOB activity was called the 1818Fproduction efficiency, ?(EOB), which reflects mainly the
irradiation condition and the transfer loss. The ratio of the measured EOS activity to the
18measured EOB activity corrected the [F]FDG synthesis efficiency, ?(SYN). Fig. 1 shows
the ?(EOB) for each batch of operation. The average value of ?(EOB) was 0.76?0.07. The fluctuation of ?(EOB) due to irradiation conditions, such as beam focus, heat transfer, etc. was relatively smaller. For some batches of operation, where ?(EOB) was quite low, different kinds of transfer problems, such as, leakage and filer blockage had been identified. Fig.2 shows the ?(SYN) for the two synthesizers alternatively used in the Shin Kong Cyclotron Center. The average of ?(SYN) was 69?7.4% and 53?15.9% for Coincidence and Microlab2, respectively. It should be noticed that the chemical synthesis following the 18Fproduction is a separate issue and cannot be correlated with irradiation conditions. It involves fast chemistry at a no-carrier-added level and the yield may due to several reasons. The synthesizer performance of Coincidence was superior to that of Microlab2 in terms of efficiency, stability, and synthesis time, which was 28 min for the former and 35min for the latter. Therefore, Shin Kong Cyclotron Center planed to replace Microlab2 with another Coincidence.
3.2 Neutron counting rate
The moderated 3He neutron detector was installed near the self-shielded MINItrace
TMcyclotron. As expected, the measured neutron counting rate depends on the
position of the target.
Target Sites 1 and 3 were used alternatively in the routine operations of the Shin Kong Cyclotron Center.
For normal operations using a 30?A beam current, the typical neutron counting rates were measured about 13000 and 30000 counts per minute(cpm) for Target Sites 1 and 3, respectively. For truly correlating the 18Fproduction rate and the neutron counting rate, neutrons not coming from the 18O(p,n)18F reaction must be subtracted. Since the
1818threshold energy of the O(p,n)F reaction is much higher than 9.6Mev, pure water was
18used to replace the enriched [O]water to obtain the background. It contributed only
about 10% to the total neutron counts. Usually, the neutron counting rate kept quite stable during beam irradiation; the fluctuation is in the level of about 1.5～3.5% depending on the target positions. However, an abnormal phenomenon was recorded on 12/08/2003 as shown in Fig.3 that shows a great advantage of installing a neutron monitor. The beam current was kept at 30?A during irradiation, short beam trips occurred several times due to the discharge in the negative electrode and affected neutron counting rates accordingly. It is interesting to note that after the time stamp of 7.5-h the neutron counting rate decreased gradually from 13000 to 2500cpm. It was because of the failure in the seal foil
18of the target body that caused the leak of Owater. But the operator was not aware of
this failure and still kept beam on still EOB. It may causes serious component damage to the target chamber but could be avoided by using an on-line neutron monitor.
There were several leakage phenomena observed in about 850 neutron monitoring records. Most of them were caused due to the looseness of the delivery tube joints.
Fig.4 shows the history record of the average net neutron counting rate per unit current for the two target sites. There were two target bodies used alternatively for each target site.
The average net neutron counting rate at Target Site 3 was about twice higher than that at Target Site 1 which was due to the different beam direction relative to the moderated
3Hedetector and the different shielding thickness traversed by neutrons from the target to the detector. At each target site the average net neutron counting rate shows two peaks that may be attributed to the different target bodies applied. The fluctuation of the average counting rate represents the variation of the irradiation condition. The standard deviations of the net neutron counting rate as shown in Fig.4 for specific target body were found to be 2.2% and 1.4% at Target Site 1and 3, respectively.
Theoretically, the neutron counting rate multiplied by the detector efficiency should equal to the reaction rate R which could be calculated from
activity and given Eq.(2) for a known EOB TEOB. A conversion factor k was introduced as follows: R?(EOB)?
K=NaveNave[1?exp(??TEOB)] (3)
Where Nave is the average net neutron counting rate calculated from each batch. For the calculation of k the measured EOB activity at the entrance of the synthesizer was adopted. Fig.5 shows the conversion factor k for Target Site 1 and 3. The standard deviation of the conversion factor k was found to be 6.9% and 8.7% for Target Sites 1 and 3, respectively. Since both the measured EOB activity and the neutron counting rate were both affected
by the irradiation condition, the fluctuation of the conversion factor k would reflect the