Development of a compact multiprobe system for monitoring positron-emitting tracers in plant stems

Nondestructive monitoring of positron-emitting tracers in plant bodies at multiple points, including points separated by large distances, has been realised with the positron multiprobe system (PMPS) to investigate graminaceous plants, whose stems and leaves have simple shapes. Recently, the translocation of photosynthates into fruits has been studied intensively using a 11C tracer. The relatively complex shapes of the stems and leaves of these plants sometimes prevent the detector heads of the PMPS from approaching and being fixed to the target stem properly because of its relatively large and heavy detector heads based on photomultiplier tubes. Owing to the compactness, lightweightness and recent advances of silicon photomultipliers (SiPMs), fabricating compact and lightweight detector heads has become possible. In this study, we developed a compact PMPS (CPMPS) using SiPMs and successfully demonstrated its capability for monitoring a 11C tracer in strawberry stems. Moreover, we found that energy-window filtering markedly reduced noise events without radiation shielding. The dominant ionisation events detected by the CPMPS were Compton scattering and subsequent photoabsorption of a single 511 keV gamma ray, suggesting that the ionisation events of single-gamma-ray emitters, such as 42K, 43K, 54Mn, 59Fe and 65Zn, can be detected by the CPMPS. The developed CPMPS can also be applied to study the physiology of other plants with intricately shaped stems and leaves, such as the tomato and eggplant.


Introduction
Nondestructive monitoring of radioactive tracers at multiple points in plant stems is essential for understanding how plants absorb and use the various elements present in the soil and atmosphere [1].In particlular, nondestructive monitoring of positron-emitting tracers in plant bodies has been realised with the positron multiprobe system (PMPS) [2] as well as positron imaging systems [1].Positron imaging systems, such as positron emission tomography systems [3,4] and planar positron imaging systems [5], are powerful tools for analysing positron-emitting-tracer distribution and movement in a given field of view (FOV).However, the FOV of these positron imaging systems sometimes cannot cover the whole plant body.In contrast, the PMPS can monitor positron-emitting tracers in plant bodies at multiple points, including points separated by large distances.The only existing PMPS, which we call "old PMPS" here, has relatively large and heavy detector heads (20 × 20 × 72 mm, 60 g) owing to the use of photomultiplier tubes (PMTs).Therefore, the old PMPS is restricted to investigations of graminaceous plants, such as rice [2,6,7] and barley [8][9][10], whose stems and leaves have simple shapes.
Recently, the translocation of photosynthates into fruits, such as the strawberry [11], tomato [12,13] and eggplant [14], has been studied intensively using a 11 C tracer.Because the detector heads are placed 180 • apart with stem sandwich (detector-stem-detector) to detect the two annihilation gamma rays in coincidence, the relatively complex shapes of the stems and leaves of these plants sometimes prevent the bulky and heavy detector heads of the old PMPS, which require support clamps fixed on a lab stand to avoid a bending stress of the stems, from approaching and being fixed to the target stem properly.This limitation is essential in plant experiments, and thus compact and lightweight detector heads improve flexibility of setups drastically, which realises nondestructive monitoring of positron-emitting tracers in plant bodies with various size and shape at given multiple points, including points separated by large distances.Owing to the compactness, lightweightness and recent advances of silicon photomultipliers (SiPMs) [15], fabricating compact and lightweight detector heads has become possible, which enables miniaturizing setups around the plant stems without support clamps.In this study, we developed a compact PMPS (CPMPS) using SiPMs and performed a demonstration experiment, which involved monitoring a 11 C tracer in strawberry stems.

Compact positron multiprobe system (CPMPS)
Table 1 shows the main specifications of the old PMPS and CPMPS.The CPMPS has 16 detector heads.Each detector head consists of a 10 mm cubic high-resolution gadolinium aluminium gallium garnet (HR-GAGG) scintillator (C&A Corp., Sendai, Japan) directly coupled to a SiPM (S14160-6050HS, Hamamatsu Photonics K. K., Hamamatsu, Japan) without any optical grease or cement.SiPMs have advantageous features such as compactness, wide spectral response range, high efficiency and high gain, while HR-GAGG scintillators are characterised by negligible hygroscopicity, high light yield and high energy resolution.
Figure 1 shows a schematic diagram of the cross-sectional view of a detector head of the CPMPS.Each HR-GAGG scintillator is supported by cushion tape and is packaged in an aluminium case.The size and weight of each detector head are 15 × 15 × 16 mm and 12 g, respectively.Figure 2 shows a schematic diagram of the CPMPS.Signals from eight detector heads are passed through voltage suppliers and then fed to preamplifiers (AD8056, Analog Devices, Wilmington, MA, U.S.A.) through 4 m coaxial cables.The transmission loss of the cable is only 2 dB at 100 MHz, which is acceptable for this system.The signals from the preamplifiers are fed to a weighted summing amplifier [16,17] to reduce the number of signals.The signals from the other eight detector heads are also fed to preamplifiers and then to the other weighted summing amplifier.The analogue signals from the weighted summing amplifiers are fed to the 100 MHz free-running analogue-to-digital converters (ADCs).The converted digital signals are integrated to calculate the raw position (i.e.detector assignment) and energy data in the field-programmable gate array (FPGA).
-2 -  1 Photon detection efficiency at each emission wavelength considering each effective area.
The operating voltage on each SiPM is supplied by the voltage supplier.Each signal amplitude at the voltage supplier is approximately 300 mV for an energy deposition of 662 keV, which corresponds to approximately 30,000 emitted scintillation photons, as shown in table 1.Because each SiPM has the temperature coefficient of the recommended reverse voltage, 34 mV/ • C, the operating voltage on each SiPM is adjusted automatically using the temperature sensor shown in figure 1.The circuit configuration of the voltage adjustment is the same as that of the power supply (C14156, Hamamatsu Photonics K. K.).The energy data of each detector head are calibrated at 511 keV, and each calibration curve is assumed linear.The energy resolution of each detector head is 13%-14% in full width at half maximum at 511 keV.
Because the clock frequency of the FPGA is 100 MHz, the FPGA detects coincidences of the digital signals from the two ADCs in a variable time window from ±10 to ±160 ns.The coincidence time window is ±50 ns, which was determined using a process similar to that in ref. [18].First, the 22 Na source was measured for 10 s with each changing coincidence time window.Second, the dependence of the coincidence count rate on the coincidence time window was plotted and fitted with an error function.Finally, the minimum coincidence time window where the coincidence count rate exceeded the plateau of the fitting curve within the standard deviation was determined to be ±50 ns.The ADCs and FPGA used in the CPMPS are the same as those used in ref. [19].

Plant material and growth conditions
A strawberry plant (Fragaria × ananassa Duch.) was used in this study.Detailed descriptions of the plant growth conditions can be found in refs.[4,11,20].Briefly, the plants were grown in plastic pots filled with a substrate consisting of peat moss, coconut shells and charcoal [3:5:2 (v/v/v)].The pots were placed in a plant growth chamber at the Takasaki Institute for Advanced Quantum Science, National Institutes for Quantum Science and Technology (QST), Gunma, Japan (36 • 18.1 ′ N, 139 • 04.4 ′ E).The growth chamber settings were as follows: photosynthetically active radiation (PAR) using light-emitting diode lamps of 400 μmol m −2 s −1 , air temperature of 20 • C, relative humidity of 60%, CO 2 concentration of 380 μmol mol −1 and day:dark photoperiod of 12:12 h.Plants were hand watered with deionised water as needed.A strawberry plant with one primary fruit, one secondary fruit and three tertiary fruits was used in the experiment.

Tracer production and feeding
A 11 C-labelled carbon dioxide ( 11 CO 2 ) tracer was produced by bombarding pure nitrogen gas with 10 MeV protons from the cyclotron at Takasaki Institute for Advanced Quantum Science, QST, Gunma, Japan [11,12,21].The irradiated gas containing nitrogen and 11 CO 2 passed through a stainless-steel trap immersed in liquid nitrogen, and thus only 11 CO 2 was collected as dry ice.The resultant 176 MBq of 11 CO 2 was collected and fed to a leaf of the plant with air.

JINST 19 P04023
The test plant at the third inflorescence fruiting stage consisted of one primary fruit, two secondary fruits and three tertiary fruits and had a cymose inflorescence structure.Here, one of two secondary fruits was picked the day before the 11 CO 2 -tracer feeding (upper, figure 3).To demonstrate the adequate performance of the CPMPS, the translocation routes of the photosynthetically fixed carbon were controlled in the rest of the plant.Prior to 11 CO 2 -tracer feeding, the plant leaf was enclosed and fixed between two acrylic plates (inner space of 20 cm in length, 15 cm in width and 1 cm in depth).The edge of the acrylic plates sandwiching the leaf was sealed with an O-ring.In addition, the gap around the petiole insertion hole of the acrylic plates was filled with plastic clay to prevent 11 CO 2 leakage.Then, ambient air was flowed through the chamber to keep the leaf humid.The 11 CO 2 was administered to the leaf in the flow-through acrylic chamber within a 1 min pulse under continuous air flow of 300 mL min −1 and PAR of 400 μmol m −2 s −1 for 20 min. 11CO 2 exiting the other end of the chamber was collected with soda lime (Wako Pure Chemical Industries, Osaka, Japan), a CO 2 absorber, in an acrylic container connected to the chamber.Leaf fixation of 158 MBq was determined by subtracting the radioactivity of exited 11 CO 2 absorbed by soda lime from that of injected 11 CO 2 .After 11 CO 2 -tracer feeding for 20 min, the test plant was detached from the flow-through acrylic chamber and immediately transferred to an experimental chamber.were assigned and fixed to plant stems with dental spacer material (DF-TAK003, Takara Belmont Corp., Osaka, Japan).The distance between the two detector heads of each pair was approximately 10 mm.The measurement started 29 min after tracer feeding and continued for 164 min, which is similar to the duration of other 11 CO 2 -tracer experiments using strawberry plants [4,11,20].The environmental conditions (PAR, air temperature and relative humidity) in the experimental chamber was the same as in the growth chamber.

Results
Figure 4 shows a two-dimensional energy spectrum of the data collected by a pair of detector heads (Nos. 1 and 9) as an example.The upper graph of figure 5 shows the time profiles of the count rates corrected by the half-life of 11 C (20.34 min [22]).The lower graph of figure 5 shows the time profiles of the corrected count rates filtered by the energy window in which the sum of the deposited energies was greater than 600 keV and each deposited energy was less than 600 keV.The former condition of the energy window was imposed to eliminate the coincidence events of Compton scattering and subsequent photoabsorption of a single 511 keV gamma ray, considering energy resolution.The latter was imposed to eliminate part of the ionisation events of two (or more) energy depositions in a single scintillator by two (or more) 511 keV gamma rays, considering energy resolution.

Discussion
As shown figure 5, the corrected count rates in 2 pairs of detector heads, Nos. 1 and 9 and Nos. 3 and 11, gradually increased and started to plateau at approximately 6000 s.This result is consistent -5 -

Fruits
No. 10 No. 12 No. 9 As shown in figure 4, the dominant ionisation events were the coincidence events of Compton scattering and subsequent photoabsorption of a single 511 keV gamma ray rather than photoabsorption, Compton scattering or both of a pair of annihilation 511 keV gamma rays.These dominant ionisation events contained both signal events (from the stem between the two detector heads) and noise events (from the other parts of the plant, especially the leaf to which the tracer was fed), as shown in the upper graph of figure 5.As shown in the lower graph of figure 5, the energy window eliminated almost all the noise events (and part of the signal events) without radiation shielding, and the movement of the tracer to the largest fruit (group) was observed clearly.However, the corrected count rate in detector pair Nos. 1 and 9 differed from that in detector pair Nos. 3 and 11, despite the almost-zero count rates in detector pair Nos. 2 and 10 and detector pair Nos. 4 and 12.This was attributed to not only the effect of plant physiology but also the difference in sensitivity between the detector pairs arising from the detector setup (mainly the distance between the two detector heads).Indeed, a 1 mm error of the distance between the emission point and the scintillator surface results in an approximately 20% change in the solid angle subtended at the emission point by the scintillator surface, which approximately corresponds to a 20% change in the count rate.A method (or jig) to achieve precise detector setup will improve the difference, and its development will be the focus of a future study.
-7 - (Lower) Time profiles of the corrected count rates filtered by the energy window in which the sum of the deposited energies was greater than 600 keV and each deposited energy was less than 600 keV.The corrected count rates of the 4 pairs of detector heads assigned to plant stems, Nos. 1 and 9, Nos. 2 and 10, Nos.-8 -

Conclusions
We developed the CPMPS using SiPMs and successfully demonstrated its capability for monitoring a 11 C tracer in strawberry stems.Moreover, we found that energy-window filtering markedly reduced noise events without radiation shielding.Each of the dominant ionisation events seemed to be produced by a single 511 keV gamma ray.This suggests that single-gamma-ray emitters, such as 42 K, 43 K, 54 Mn, 59 Fe and 65 Zn [23], can be monitored using the CPMPS, although they would generate considerable noise events.Thus, applying the CPMPS to monitor these single-gamma-ray emitters is challenging and will require further studies.Future research plans include developing a method to achieve precise detector setup and applying the CPMPS to monitor single-gamma-ray emitters as well as to investigate other plants, such as the tomato [12,13] and eggplant [14].

Figure 1 .
Figure 1.Schematic diagram of the cross-sectional view of a detector head of the CPMPS.

Figure 2 .
Figure 2. Schematic diagram of the CPMPS.Arrows represent signal cables.

Figure 3 (
Figure 3 (lower left) shows a photograph of some of the detector heads and figure 3 (lower right) shows a photograph of an experimental setup.Four pairs of detector heads (Nos.1-4 and 9-12)were assigned and fixed to plant stems with dental spacer material (DF-TAK003, Takara Belmont Corp., Osaka, Japan).The distance between the two detector heads of each pair was approximately 10 mm.The measurement started 29 min after tracer feeding and continued for 164 min, which is similar to the duration of other 11 CO 2 -tracer experiments using strawberry plants[4,11,20].The environmental conditions (PAR, air temperature and relative humidity) in the experimental chamber was the same as in the growth chamber.

Figure 3 .Figure 4 .
Figure 3. (Upper) Schematic diagram of the experimental setup.The fruit marked by dotted lines was picked the day before tracer feeding.(Lower left) Photograph of some of the detector heads of the CPMPS.(Lower right) Photograph of an experimental setup.The four pairs of detector heads assigned to plant stems are marked with red open circles.

Figure 5 .
Figure 5. (Upper) Time profiles of the count rates corrected by the half-life of 11 C (corrected count rates).(Lower)Time profiles of the corrected count rates filtered by the energy window in which the sum of the deposited energies was greater than 600 keV and each deposited energy was less than 600 keV.The corrected count rates of the 4 pairs of detector heads assigned to plant stems, Nos. 1 and 9, Nos. 2 and 10, Nos. 3 and 11 and Nos. 4 and 12, are displayed as black closed circles, red closed squares, green closed triangles and blue closed reversed triangles, respectively.Error bars represent the standard deviation of a Poisson distribution.
Figure 5. (Upper) Time profiles of the count rates corrected by the half-life of 11 C (corrected count rates).(Lower)Time profiles of the corrected count rates filtered by the energy window in which the sum of the deposited energies was greater than 600 keV and each deposited energy was less than 600 keV.The corrected count rates of the 4 pairs of detector heads assigned to plant stems, Nos. 1 and 9, Nos. 2 and 10, Nos. 3 and 11 and Nos. 4 and 12, are displayed as black closed circles, red closed squares, green closed triangles and blue closed reversed triangles, respectively.Error bars represent the standard deviation of a Poisson distribution.