- Email: [email protected]
Development and characterization of a Drop-on-Demand inkjet printing system for nuclear target fabrication
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49
Contents lists available at ScienceDirect
Nuclear Inst. and Methods in Physics Research, A journal homepage: www.elsevier.com/locate/nima
Development and characterization of a Drop-on-Demand inkjet printing system for nuclear target fabrication R. Haas a,b, *, S. Lohse a,b , Ch.E. Düllmann a,b,c , K. Eberhardt a,b , C. Mokry a,b , J. Runke a,c a b c
Johannes Gutenberg-Universität Mainz, Institut für Kernchemie, Fritz Strassmann Weg 2, 55128 Mainz, Germany Helmholtz-Institut Mainz, Staudinger Weg 18, 55128 Mainz, Germany GSI Helmholtzzentrum für Schwerionenforschung GmbH, Planckstraße 1, 64291 Darmstadt, Germany
a r t i c l e
i n f o
Keywords: Drop-on-Demand Inkjet printing Thin layers Storage phosphor imaging plates Radiographic imaging Targets
a b s t r a c t A novel target preparation method based on Drop-on-Demand (DoD) inkjet printing has been developed. Conventional preparation methods like the electrochemical method ‘‘Molecular Plating’’ or the ‘‘Polymer-Assisted Deposition Method’’ are often limited, e.g., concerning the dimensions and geometries of depositions or by the requirement for electrically conducting substrates. Here, we report on the development of a new technique, which overcomes such limits by using a commercially available DoD dispenser. A variety of solutions with volumes down to 5 nL can be dispensed onto every manageable substrate. The dispensed volumes were determined with a radioactive tracer and the deposits of evaporated salt solutions were investigated on titanium and graphene foils. Additionally, the high precision of the printing system with which individual drops can be positioned was used to determine the spatial resolution of storage phosphor imaging plates with three tracers of different 𝛽-decay energies. The new technique is able to produce new kinds of targets with improved spatial geometries and thin layer deposits. © 2017 Elsevier B.V. All rights reserved.
1. Introduction There is a high demand for well prepared thin actinide layers serving as targets in both nuclear chemistry and nuclear physics applications. These include accelerator-based experiments like the production of the heaviest elements [1], nuclear reaction data needs from basic and applied sciences [2,3], or sources of fission fragments [4] as well as 𝛼daughter recoil nuclei [5]. Generally, targets have to meet high demands in terms of chemical purity, homogeneity, thickness, as well as adherence to the deposition substrate. One of the most successful production methods for such thin layer depositions for nuclear applications is the electrochemical deposition from organic solvents called Molecular Plating (MP) [6,7]. Alternative methods include manual pipetting on superhydrophobic surfaces [8] or polymer assisted deposition (PAD) [9,10], developed by Jia et al. [11], the latter being suitable for the preparation of thin films of metal oxides. Naturally, all methods have some inherent limitations, like the need for an electrically conductive substrate in MP, or the limited size and constrained geometric form the target can take in MP and PAD. Manual pipetting, on the other hand, suffers from limited accuracy. An automated pipetting system has recently been built by a group at the Paul Scherrer Institute (PSI), Villigen, Switzerland, to
overcome the limits in accuracy of manual pipetting and has been used for the production of targets for experiments of the n_TOF collaboration at CERN [12]. The application of precision printing systems like DoD inkjet printing has recently spread increasingly in many branches of the natural sciences. The DoD technique is widely used in commercial printers for producing droplets of small, well defined volumes and to dispense them in a defined area, e.g., for barcode labeling. Here, single droplet dosage is performed either piezoelectrically or thermally [13]. By using modified commercially available inkjet printers, the thermal DoD method and its produced deposits were investigated by Fittschen et al. [14]. Thermal DoD printers are able to produce droplets of solutions in the range of picoliters with a good reproducibility, which lead to deposits with diameters of 5 μm to 20 μm in a hemispherical shape. In contrast, simple droplet evaporation of larger droplets is known to produce ringshaped residues [14], which can best be overcome by depositing on superhydrophobic surfaces [8], which, though, introduces new limitations. For radiochemical applications, thermal DoD printers have a big disadvantage because a part of the solution is volatilized inside the printing head and the internal reservoir gets contaminated and cannot be
* Corresponding author at: Johannes Gutenberg-Universität Mainz, Institut für Kernchemie, Fritz Strassmann Weg 2, 55128 Mainz, Germany.
E-mail address: [email protected] (R. Haas). http://dx.doi.org/10.1016/j.nima.2017.08.027 Received 29 June 2017; Received in revised form 14 July 2017; Accepted 15 August 2017 Available online 24 August 2017 0168-9002/© 2017 Elsevier B.V. All rights reserved.
R. Haas et al.
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49
exchanged. Hence, the printing head would have to be fully exchanged for a different application to avoid cross-contamination effects. As an alternative to thermal inkjet printing piezoelectric DoD dispensers have been investigated. One such system is the PipeJet® P9 Nanodispenser from Biofluidix. This features an external reservoir, which is in contrast to thermal inkjet printers. It was shown that the modified thermal inkjet printer has a smaller deposition volume as well as the ability to print bands, whereas the piezoelectric dispenser has a slightly better precision in printing spots [15]. Piezoelectric DoD printers are currently used, e.g., in energy research to produce inkjet-printed Cu2 ZnSn(S, Se)4 solar cells [16]. Piezoelectric dispensers have an external reservoir, which is easily exchangeable and the solution is not in contact with internal parts of the dispenser. Therefore, in this work, a new method to produce targets based on piezoelectric DoD inkjet printing has been developed and basic features of the produced thin layer deposits have been characterized. The setup of the system is built up of a sample stage and a dispenser based on the DoD technology with an external reservoir for liquids. It is possible to dispense liquids over a wide range of viscosities, from water-based solutions to, e.g., glycerin-based ones. Droplet volumes between 5 nL and 60 nL can be produced, which is far below the limit of commercially available manual pipettes. In this work, different droplet volumes for water-based solutions with a specific concentration were verified using radioactive tracers. Additionally, the residue after droplet evaporation was characterized via optical microscopy as well as scanning electron microscopy (SEM). The system was finally applied to produce well defined patterns of radioactive species to study the spatial resolution of an autoradiographic imaging system as a function of 𝛽-decay energy.
Fig. 1. View of the DoD printing system containing the piezo dispenser (1) and two compact motorized translation stages (2) with a holder for substrates with a diameter of 26 mm (3).
2. DoD inkjet printing system 2.1. Setup of the printing system The inkjet printing system comprises a DoD dispenser (Biofluidix, PipeJet® P9 Nanodispenser) for the droplet production, and two compact motorized translation stages (Thorlabs) to move a substrate in two dimensions, see Fig. 1. The dispenser consists of a printing head with an electronically controlled piezo-driven piston inside its casing and a replaceable tip connected to an external reservoir. The tips are made of polyimide and polypropylene and have an inner diameter of 200 μm. To connect the tip and the reservoir, an 8 cm-long flexible tube (Proliquid GmbH, LMT-55) made from TYGON® is used. The reservoir is a commercially available pipette tip made of polypropylene and can be fixed to the holder of the dispenser. To adjust the height of the dispenser over the substrate, the former is mounted to a stand. Each translation stage has a travel range of 50 mm and a bidirectional repeatability of 1.6 μm [17]. To fix substrates in a well-defined position during movements, a holder for circular substrates with a diameter of 26 mm is mounted on the top stage. This can be easily replaced with holders for targets of different geometries, providing for the flexibility of the setup. The liquid is held in external, easily exchangeable and cheap parts and not in the casing of the dispenser, which is a significant advantage over normal inkjet cartridges and especially relevant for applications using a multitude of sample solutions. Thus the work with radioactive substances is simplified, because the contaminated parts can easily be exchanged. For good printing results, a vertical distance of 1 mm to 2 mm between the tip and the substrate is recommended by the manufacturer. Liquids and solutions with a viscosity of 0.5 mPa⋅s to 500 mPa⋅s and a surface tension of 30 mN/m to 76 mN/m are suitable for dispensing. The manufacturer specifies a volume precision of <3% and accuracy of <10% for the dispenser [18]. Drop volumes of 5 nL to 60 nL can be generated in dependence of the used tip diameter. For each drop volume, concentration of a solution and solvent, the stroke velocity of the piston has to be calibrated via a manufacturer-delivered software. An optimal value is found when the dispenser can continuously print without
Fig. 2. Schematic representation of the dispensing process [18]. A liquid is kept inside the elastic polymer tip by its surface tension until a fast displacement of the piezo driven piston. The liquid is subsequently pressed to both directions, and a droplet forms at the lower end of the tip and is released. After a slow release of the piston, new liquid runs from the reservoir into the tip to fill up the tip again.
generating satellite drops. The drop volume is set by the stroke range and the stroke velocity of the piezo-driven piston inside the dispenser. The schematic principle of the dispenser is illustrated in Fig. 2. Once filled, the liquid is kept inside the elastic polymer tip due to its surface tension. After a fast displacement of the piston with a specific stroke range and stroke velocity, the liquid is forced out of the respective part of the tip hit by the piston in both directions. At the bottom of the tip, a drop is generated and falls off. By a slow release, new liquid from the reservoir fills up the tip again. The stroke range influences the suppressed liquid volume inside the tip and hence is crucial for setting a drop volume. Additionally, the stroke velocity enables dispensing of a specific volume. If the stroke velocity is set too low, the kinetic energy is insufficient to let the formed drop fall off the tip. Too high values cause spraying or the generation of satellite drops. Therefore, the stroke velocity has to be calibrated for each solution and desired volume. Typical stroke velocity values are 60 μm∕ms to 100 μm∕ms for water-based samples, 50 μm∕ms to 80 μm∕ms for organic solvents and >150 μm∕ms for viscous media. The maximum frequency of the dispenser to produce droplets lies at 100 Hz depending on the liquid used. After the dispensing process the drops evaporate on the substrate, leaving behind the residue of the dissolved species. 44
R. Haas et al.
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49
were performed with stroke velocities of 100 μm∕ms and 145 μm∕ms. To obtain enough activity with a single test tube for gamma spectroscopy, 100 drops were dispensed in each of the test tubes. Additionally, 1 μL of the radioactive solution was dispensed into a third test tube with a pipette (Eppendorf, reference 2, 0.5 μL to 10 μL) as a standard volume. Every test tube was filled with isopropanol to 1 mL to obtain a similar sample geometry, and was measured with a gamma detector (Canberra, GX6020, rel. eff. 64.8%) for 30 min. The results obtained with 140 La were verified in a second identical series of measurements, where the 198g Au tracer was used. For this 500 μL of the 198g Au solution was filled into the reservoir of the dispenser and different volumes and velocities were dispensed under the same conditions as the 140 La tracer. To eliminate any influence of the set stroke velocity for different volumes, the stroke velocity was kept fixed at 90 μm∕ms. Every sample was measured for 10 min with the gamma detector used before. Scanning electron microscopy (Philips Inc., XL 30) was used to investigate the distribution pattern of the deposits. For this, an unirradiated solution of the La(NO3 )3 was dispensed in a rectangular pattern onto a titanium foil and a graphene foil. The settings of the dispenser were 8 nL volume with a stroke velocity of 100 μm∕ms. The sequence was a 10 × 10 pattern of points with an interval of 0.5 mm.
Fig. 3. Example of a planned sequence preview with equidistant drops in a circular area with 5 mm diameter.
2.2. Software To ensure a fully automated process of printing, including the synchronization of the movement of the stages and the timing of the dispenser, a custom software suite based on LabView has been developed. It uses a two-axis coordinate system with a grid resolution of 0.01 mm and can support up to 64000 individual drops. A series of coordinates is referred to as a sequence and represents all information needed to print a target. Sequences can be exported to files for later use and sharing. The planning of sequences can also be done independently of the printer by a standalone software ‘‘planning tool’’ to increase efficient usage of the actual printer hardware. The possibilities to use polar coordinates or to create repeatable patterns of coordinates are also integrated in the planning tool. Furthermore, the software suite provides the user with a preview of the planned sequence and is able to simulate the printing process for testing and controlling purposes. An example of a planned sequence preview consisting of equidistant drops, cut to fit a circular 5 mm target, is given in Fig. 3. A further function is the conversion of graphical coordinates into Cartesian coordinations, which allows the user to transform a monochrome bitmap into a sequence.
2.4. Results and discussion 2.4.1. Drop volume The results of the determination of the drop volume using 140 La tracer are shown in Table 1 and give a precision of (5.0 ± 2.5)% for a volume of 8 nL. The results of the 198g Au tracer verify the precision of the 140 La tracer with an overall precision of (5.2 ± 1.4)%. The influence of the stroke velocity on the drop volume is clearly visible in the data given in Table 1. The deviation of the drop volume of nearly 15% for three different stroke velocities affirms the importance of a calibration for a selected stroke velocity before printing. Otherwise, the dispensed drop volume may be up to 40% higher than the preset volume (cf. Table 1). The volume also depends on the viscosity of the used solution, which depends on the concentration of the dissolved salt and the used solvent. The average deviation of three measured samples for each drop volume is 1.5%, which confirms a good reproducibility, once the calibration is performed. This is essential to produce samples containing a well-defined amount of deposited material.
2.3. Production and characterization of DoD targets 2.4.2. Deposition dimensions on Ti foils and on graphene SEM pictures of lanthanum nitrate deposits are shown in Fig. 4. They are visible by the material contrast of the SEM pictures. Thus, the deposits appear darker than the titanium substrate (a) and brighter than the graphene substrate (b). While the deposits on titanium (a) are irregular and have diameters of about 400 μm, the ones on graphene (b) have a smoother and more regular surface and have much smaller diameters of about 170 μm. The structure of the titanium foil is visible through the deposit in (a), indicating this to be very thin. The thickness of the deposits can be determined from the concentration of the solution, the dispensed drop volume and the diameter of the deposit. The deposits on Ti foil have an average areal density of about 16 μg⋅cm-2 , and those on graphene foil about 88 μg⋅cm-2 . On graphene, they seem to have a hemispherical shape, as follows from the contrast in height within the deposition, which is seen in addition to the material contrast of the SEM image. This shape and the smaller diameter result from different surface properties. The graphene foils have a more hydrophobic surface than the untreated titanium foils [19]. This property causes a higher contact angle between droplets and the substrate, which has already been investigated for the preparation of targets on superhydrophobic surfaces [8]. By this, the generation of salt rings and other defects can be avoided during evaporation of the droplets.
2.3.1. Reagents and materials 140 La (E = 1.4 MeV, E = 1596 keV, t 198g Au 𝛽 𝛾 1∕2 = 40.272 h) and (E𝛽 = 1.0 MeV, E𝛾 = 412 keV, t1∕2 = 2.69 d) were used as radioactive tracers. A solution was prepared by dissolving lanthanum nitrate (natural isotopic composition) in 0.1 mol/L nitric acid to a concentration of 2.5 mg/mL with regard to lanthanum. The radioactive tracer 140 La was produced by a 6 h irradiation of 1 mL of the La(NO3 )3 solution with thermal neutrons (flux 0.7 × 1012 n⋅cm-2 ⋅s-1 ) in the research reactor TRIGA Mainz. 198g Au was produced by irradiation of an ICPMS standard solution (Merck) of 1 mg/mL Au in 10% HCl for 6 h with the same conditions like 140 La. Circular polycrystalline graphene foils (thickness 12 μm) used as substrates for the investigation of the deposits were purchased from Applied Nanotech Inc. and had a diameter of 26 mm. Titanium foils (thickness 50 μm) with a diameter of 25 mm were also used as substrates and were washed with isopropanol before usage. 2.3.2. Instrumentation and methods To determine the dispensed volume, a small test tube made from poly(methyl methacrylate) (PMMA) was fixed under the dispenser. The reservoir of the dispenser was filled with 500 μL of the radioactive 140 La solution. The drop volume was set to 8 nL. Two measurements 45
R. Haas et al.
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49 Table 1 Analysis of the dispensed drop volume for different tracers, set volumes and stroke velocities. Tracer
concentration (g/L)
Set drop volume (nL)
Stroke velocity (μm/ms)
Measured drop volume (nL)
140 La
2.5 2.5 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
8 8 8 8 8 10 12 15 17 20
100 145 70 90 110 90 90 90 90 90
8.4 ± 0.2 8.7 ± 0.9 8.5 ± 0.1 9.6 ± 0.3 11.4 ± 0.1 10.3 ± 0.4 12.7 ± 0.1 15.8 ± 0.1 17.5 ± 0.1 21.6 ± 0.2
140 La 198g Au 198g
Au Au 198g Au 198g Au 198g Au 198g Au 198g Au 198g
These plates consist of a layer of BaFBr:Eu2+ crystals with grain sizes of about 5 μm as well as a supportive and a protective layer [20]. By exposure to a radioactive sample, the electrons of the crystals get trapped in an empty lattice of the F or Br ion. After a scan with a 650 nm HeNe-Laser the spatial information can be read out by phosphorescence of the luminescence center Eu2+ [21]. The minimum resolution of these plates depends on the chosen readout pixel size. For198g Au (E𝛽 = 1.0 MeV, E𝛾 = 412 keV) a spatial resolution below 200 μm was found by measuring a radioactive gold wire with a thickness of 2.5 μm [22]. In the present publication the resolution was determined by a different technique. For this, patterns of tracers with different 𝛽-decay energies were printed onto graphene foils with the developed DoD printing system. This technique to acquire the spatial resolution was used last by Johnston et al. with 32 P, 14 C and 35 S on older storage phosphor plates [23]. The limits of the spatial resolution were set at the Rayleigh criterion, when the saddle point between two spots is at 8/𝜋 2 ≈ 80% of the maximum intensity, and at 50% of the maximum intensity. The spatial resolution of the actual storage phosphor plates is determined by the same principle. 3.2. Experimental 3.2.1. Reagents and materials To obtain a most homogeneous deposition with every drop, circular graphene foils (thickness 12 μm, Applied Nanotech Inc.) were used as substrates. Three different tracers were used, namely (i) the low-energy pure 𝛽 - -emitter 35 S (E𝛽 = 0.2 MeV, t1∕2 = 87.5 d), (ii) the high-energy pure 𝛽 - -emitter 32 P (E𝛽 = 1.7 MeV, t1∕2 = 14.26 d) and (iii) the 𝛽 - /𝛾emitter 198g Au (E𝛽 = 1.0 MeV, E𝛾 = 412 keV, t1∕2 = 2.69 d). To produce the 198g Au tracer an ICP-MS standard solution (Merck) of 1 mg/mL Au in 10% HCl was used. The low-energy 𝛽-tracer 35 S was produced in the 35 Cl(n, p)35 S reaction on the large amount of chlorine by irradiation of the Au/HCl solution. It was induced by the epithermal part of the TRIGA neutron spectrum. A solution of NH4 H2 PO4 in millipore water with a concentration of 50 mg/mL with regard to phosphor was prepared to produce the 32 P tracer. 1 mL of both solutions were irradiated for 6 h in the TRIGA-Mainz research reactor. To get a solution with only 35 S as tracer, the gold solution was reused after about 40 days decay time (16 half-lifes of 198g Au), with only the long-lived 35 S-activity remaining.
Fig. 4. SEM pictures of lanthanum nitrate deposits on a titanium foil in (a) and on a graphene foil in (b). The drops had a volume of 8 nL and were printed with a stroke velocity of 100 μm∕ms.
3.2.2. Instrumentation and methods Scanning electron microscopy (Philips Inc., XL 30) was used to investigate the average diameter of the printed drops. Therefore, patterns of both unirradiated solutions were printed with 10 × 10 drops in an 0.5 mm interval onto graphene foils. The dispenser was set for a drop volume of 8 nL for both solutions and a stroke velocity of 90 μm∕ms for the 35 S tracer, 80 μm∕ms for the 198g Au tracer, and 68 μm∕ms for the 32 P tracer. These settings were retained for printings with the irradiated solutions. The irradiated solutions were printed in equidistant line patterns with increasing distances between two drops (cf. Fig. 5). The pattern consisted of rows with ten drops each and a distance of 1 mm between two rows. The smallest achievable distance between two drop centers without coalescing, as determined in separate studies with the
3. Application of the DoD system: determination of the spatial resolution of an autoradiographic imaging system 3.1. Introduction The print of patterns with different spacings enables to determine the spatial resolution of a radiographic imager (Fujifilm, Typhoon™ FLA-7000) with this system. Here, various imaging plates are available, depending on the kind of radioactive source used for imaging. Our imaging plates (Fujifilm, BAS IP SR 2040) are designed for 𝛽/𝛾-emitters. 46
R. Haas et al.
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49
achieve a good contrast. After every sample exposure the imaging plates were exposed to light with an IP Eraser 3 from Fujifilm for reutilization. The intensity scale of the resulting images was converted into a colored scale and the intensity of all rows was evaluated as height profiles. 3.3. Results and discussion Radiographic images of the 35 S (a), 32 P (b), and 198g Au (c) pattern are given in Fig. 6. The color scale represents the amount of absorbed energy and is only qualitative, as for all radiographic images. It shows even fine differences in the intensity between two deposits. Along the framed rows the intensity was evaluated as a height profile. Intensity profiles of three rows of the 198g Au tracer are given in Fig. 7 as an example. In profile 2 the saddle between two deposits is at 50% of the maximum intensity, in profile 3 it is at more than 80% of the maximum intensity. Every radiographic image was evaluated by these criteria. These limits are reached in rows 9 and 12 of (a), rows 8 and 12 of (b), and rows 7 and 11 of (c). The diameters of the deposits were measured in SEM pictures of further unirradiated patterns and were about (500 ± 16) μm for the sulfur deposition, (260 ± 10) μm for phosphor, and (374 ± 10) μm for gold. Due to the distance between two drop centers, the resulting spacing between two 35 S deposits of the measured rows is (250 ± 16) μm in (a.1), (163 ± 16) μm in (a.2), and (72 ± 16) μm in (a.3). By the same evaluation, the spacings of the 32 P deposits are (490 ± 10) μm in (b.1), (315 ± 10) μm in (b.2), and (215 ± 10) μm in (b.3) and the spacings of the 198g Au deposits are (376 ± 10) μm in (c.1), (226 ± 10) μm in (c.2), and (126 ± 10) μm in (c.3). Following the principle of Johnston et al. [23] to determine the resolution limits of the plates, the Rayleigh criteria and the 50%-limits of the investigated tracers are given in Table 2. Because the individual droplets deposits are still easily identifiable by the naked
Fig. 5. The sequence used for the determination of the spatial resolution of the storage phosphor plates.
Au solution, was 400 μm. This distance was subsequently incremented by 25 μm per each row up to a maximum of 750 μm. Radiographic Imaging was performed with a Typhoon™ FLA-7000 from Fujifilm using (Fujifilm, BAS IP SR 2040) imaging plates to investigate the spatial resolution for different radiation types and energies. To avoid any contamination of an imaging plate, a Mylar® foil was set between the sample and the plate. The exposure time was chosen in relation to the activity of the measured sample. For samples with high activity, e.g., 198g Au the exposure time was set to a few minutes whereas for samples with low activity, e.g., 32 P, it was set to up to two hours to
Fig. 6. Radiographic images of the 35 S (a), 32 P (b), and 198g Au (c) pattern on the left and REM image (d) as well as light microscope images (e, f) for comparison on the right. The sample exposure time was 2 h for 32 P, 1 h for 35 S and 5 min for 198g Au. The color scale qualitatively represents the amount of absorbed energy and is shown relative to the maximum intensity. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
47
R. Haas et al.
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49 Table 2 Limits of the spatial resolution for three tracers of different 𝛽-decay energies and the value from Ref. [22]. tracer
𝛽-decay energy (MeV)
50%-limit (μm)
Rayleigh criterion (μm)
Ref. [22] (μm)
32 P
1.7 1.0 0.2
315 ± 25 226 ± 25 163 ± 25
215 ± 25 126 ± 25 72 ± 25
– <200 –
198g Au 35 S
with higher decay energy. Considering just the 𝛽 radiation, the spatial resolution for 198g Au should be between those of 32 P and 35 S, which is the case. So the impact of the gamma radiation is not significant enough to cause a substantially lower spatial resolution for the 198g Au tracer. Also the Rayleigh criterion of 198g Au with (126±25) μm fits to the spatial resolution found by Liebe et al. of < 200 μm [22]. 4. Conclusion A novel method for the preparation of thin layer targets was developed on the basis of the Drop-on-Demand technology. A printing system was designed using a piezoelectric dispenser and a x,y-displaceable sample stage. A software was developed to enable automated printing of defined patterns. The dispense precision as well as the shape of deposits on different substrates were investigated with radioactive tracers and SEM. The dispensed drop volume is in agreement with that indicated by the manufacturer for the settings of the dispenser with an overall precision of about (5.2 ± 1.4)%, but was found to depend significantly on the used stroke velocity. On graphene foils the depositions result in nearly homogeneous and hemispherically shaped residues with diameters below 500 μm at a drop volume of 8 nL. Thus the formation of salt rings can be avoided at such small drop volumes. The thickness of the deposits can be minimized by a low concentration and large drop volumes. As a first application of the printing system, the spatial resolution of a radiographic imaging system was investigated with solutions of 32 P, 198g Au and 35 S tracers. With the high accuracy of the printing system, the spatial resolution was measured in relation to the 𝛽-decay energy, which would not easily have been achievable by manual pipetting. For the lowest energy of 0.2 MeV (35 S) the limit of the spatial resolution is at (72 ± 25) μm when defined according to the Rayleigh criterion. The developed printing system establishes a new method for precise target preparations, which overcomes limits of conventional preparation methods, with the ability of automated printing and a high accuracy. Acknowledgments The authors acknowledge the local support of the staff of the TRIGA Mainz for performing the irradiations and of the mechanical workshop at the Institute of Nuclear Chemistry. We acknowledge the financial support from the Helmholtz Institute Mainz. References [1] J. Runke, Ch.E. Düllmann, K. Eberhardt, P.A. Ellison, K.E. Gregorich, S. Hofmann, E. Jäger, B. Kindler, J.V. Kratz, J. Krier, B. Lommel, C. Mokry, H. Nitsche, J.B. Roberto, K.P. Rykaczewski, M. Schädel, P. Thörle-Pospiech, N. Trautmann, A. Yakushev, Preparation of actinide targets for the synthesis of the heaviest elements, J. Radioanal. Nucl. Chem. 299 (2014) 1081–1084. [2] B.W. Filippone, M. Wahlgren, Preparation of a 7 Be target via the molecular plating method, Nucl. Instrum. Methods A 243 (1986) 41–44. [3] A. Vascon, J. Runke, N. Trautmann, B. Cremer, K. Eberhardt, Ch.E. Düllmann, Quantitative molecular plating of large-area 242 Pu targets with improved layer properties, Appl. Radiat. Isot. 95C (2014) 36–43. [4] H.L. Adair, P.R. Kuehn, Preparation of 252 Cf neutron and fission-fragment sources, Nucl. Instrum. Methods 114 (1974) 327–332. [5] K. Eberhardt, Ch.E. Düllmann, R. Haas, C. Mokry, J. Runke, P. Thörle-Pospiech, N. Trautmann, Actinide targets for fundamental research in nuclear physics, in: Physics Procedia, Physics Procedia: 28th World Conference of the International Nuclear Target Development Society, (accepted for publication), 2017. [6] W. Parker, R. Falk, Molecular plating: A method for the electrolytic formation of thin inorganic films, Nucl. Instrum. Methods 16 (1962) 355–357.
Fig. 7. Linear profiles of the intensity along rows 1, 2 and 3 of the radiographic image of 198 Au as labeled in Fig. 6. The distances of the drop centers were (376 ± 10) μm in row 1, (226 ± 10) μm in row 2, and (126 ± 10) μm in row 3.
eye at the 50%-limit in the radiographic images, it seems more suitable to use the Rayleigh criterion as the limit for the spatial resolution of the used RI system. The results point to lower spatial resolution for tracers 48
R. Haas et al.
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49 [14] U.E.A. Fittschen, N.H. Bings, S. Hauschild, S. Forster, A.F. Kiera, E. Karavani, A. Fromsdorf, J. Thiele, G. Falkenberg, Characteristics of picoliter droplet dried residues as standards for direct analysis techniques, Anal. Chem. 80 (2008) 1967–1977. [15] G.E. Morlock, C. Oellig, L.W. Bezuidenhout, M.J. Brett, W. Schwack, Miniaturized planar chromatography using office peripherals, Anal. Chem. 82 (2010) 2940–2946. [16] X. Lin, J. Kavalakkatt, M.C. Lux-Steiner, A. Ennaoui, Inkjet-printed Cu2 ZnSn(S, Se)4 solar cells, Adv. Sci. 2 (2015) 1500028. [17] Thorlabs GmbH, MTS Series User Guide, 2016. [18] BioFluidix GmbH. PipeJet P9 Nanodispenser Datasheet, 2014. URL www.biofluidix. com. [19] K. Balasubramanian, M. Burghard, Chemie des Graphens, Chem. Unserer Zeit 45 (2011) 240–249. [20] GE Healthcare Life Sciences. Storage Phosphor Screen BAS-IP, 2012. URL https: //www.gelifesciences.com. [21] H. von Seggern, Photostimulable x-ray storage phosphors: A review of present understanding, Braz. J. Phys. 29 (1999) 254–268. [22] D. Liebe, K. Eberhardt, W. Hartmann, T. Häger, A. Hübner, J.V. Kratz, B. Kindler, B. Lommel, P. Thörle, M. Schädel, J. Steiner, The application of neutron activation analysis, scanning electron microscope, and radiographic imaging for the characterization of electrochemically deposited layers of lanthanide and actinide elements, Nucl. Instrum. Methods A 590 (2008) 145–150. [23] R.F. Johnston, S.C. Pickett, D.L. Barker, Autoradiography using storage phosphor technology, Electrophoresis 11 (1990) 355–360.
[7] A. Vascon, S. Santi, A.A. Isse, T. Reich, J. Drebert, H. Christ, Ch.E. Düllmann, K. Eberhardt, Elucidation of constant current density molecular plating, Nucl. Instrum. Methods A 696 (2012) 180–191. [8] D. Renisch, T. Beyer, K. Blaum, M. Block, Ch.E. Düllmann, K. Eberhardt, M. Eibach, S. Nagy, D. Neidherr, W. Nörtershäuser, C. Smorra, Targets on superhydrophobic surfaces for laser ablation ion sources, Nucl. Instrum. Methods A 676 (2012) 84–89. [9] M.A. Garcia, M.N. Ali, T. Parsons-Moss, P.D. Ashby, H. Nitsche, Metal oxide films produced by polymer-assisted deposition (PAD) for nuclear science applications, Thin Solid Films 516 (2008) 6261–6265. [10] M.A. Garcia, M.N. Ali, N.N. Chang, T. Parsons-Moss, P.D. Ashby, J.M. Gates, L. Stavsetra, K.E. Gregorich, H. Nitsche, Metal oxide targets produced by the polymerassisted deposition method, Nucl. Instrum. Methods A 613 (2010) 396–400. [11] Q.X. Jia, T.M. McCleskey, A.K. Burrell, Y. Lin, G.E. Collis, H. Wang, A.D.Q. Li, S.R. Foltyn, Polymer-assisted deposition of metal-oxide films, Nat. Mater. 3 (2004) 529– 532. [12] E.A. Maugeri, S. Heinitz, R. Dressler, M. Barbagallo, N. Kivel, D. Schumann, M. Ayranov, A. Musumarra, M. Gai, N. Colonna, M. Paul, S. Halfon, L. Cosentino, P. Finocchiaro, A. Pappalardo, Preparation of 7 Be targets for nuclear astrophysics research, J. Instrum. 12 (2017) P02016. [13] B. Derby, Inkjet printing of functional and structural materials: Fluid property requirements, feature stability, and resolution, Annu. Rev. Mater. Res. 40 (2010) 395–414.
49
Contents lists available at ScienceDirect
Nuclear Inst. and Methods in Physics Research, A journal homepage: www.elsevier.com/locate/nima
Development and characterization of a Drop-on-Demand inkjet printing system for nuclear target fabrication R. Haas a,b, *, S. Lohse a,b , Ch.E. Düllmann a,b,c , K. Eberhardt a,b , C. Mokry a,b , J. Runke a,c a b c
Johannes Gutenberg-Universität Mainz, Institut für Kernchemie, Fritz Strassmann Weg 2, 55128 Mainz, Germany Helmholtz-Institut Mainz, Staudinger Weg 18, 55128 Mainz, Germany GSI Helmholtzzentrum für Schwerionenforschung GmbH, Planckstraße 1, 64291 Darmstadt, Germany
a r t i c l e
i n f o
Keywords: Drop-on-Demand Inkjet printing Thin layers Storage phosphor imaging plates Radiographic imaging Targets
a b s t r a c t A novel target preparation method based on Drop-on-Demand (DoD) inkjet printing has been developed. Conventional preparation methods like the electrochemical method ‘‘Molecular Plating’’ or the ‘‘Polymer-Assisted Deposition Method’’ are often limited, e.g., concerning the dimensions and geometries of depositions or by the requirement for electrically conducting substrates. Here, we report on the development of a new technique, which overcomes such limits by using a commercially available DoD dispenser. A variety of solutions with volumes down to 5 nL can be dispensed onto every manageable substrate. The dispensed volumes were determined with a radioactive tracer and the deposits of evaporated salt solutions were investigated on titanium and graphene foils. Additionally, the high precision of the printing system with which individual drops can be positioned was used to determine the spatial resolution of storage phosphor imaging plates with three tracers of different 𝛽-decay energies. The new technique is able to produce new kinds of targets with improved spatial geometries and thin layer deposits. © 2017 Elsevier B.V. All rights reserved.
1. Introduction There is a high demand for well prepared thin actinide layers serving as targets in both nuclear chemistry and nuclear physics applications. These include accelerator-based experiments like the production of the heaviest elements [1], nuclear reaction data needs from basic and applied sciences [2,3], or sources of fission fragments [4] as well as 𝛼daughter recoil nuclei [5]. Generally, targets have to meet high demands in terms of chemical purity, homogeneity, thickness, as well as adherence to the deposition substrate. One of the most successful production methods for such thin layer depositions for nuclear applications is the electrochemical deposition from organic solvents called Molecular Plating (MP) [6,7]. Alternative methods include manual pipetting on superhydrophobic surfaces [8] or polymer assisted deposition (PAD) [9,10], developed by Jia et al. [11], the latter being suitable for the preparation of thin films of metal oxides. Naturally, all methods have some inherent limitations, like the need for an electrically conductive substrate in MP, or the limited size and constrained geometric form the target can take in MP and PAD. Manual pipetting, on the other hand, suffers from limited accuracy. An automated pipetting system has recently been built by a group at the Paul Scherrer Institute (PSI), Villigen, Switzerland, to
overcome the limits in accuracy of manual pipetting and has been used for the production of targets for experiments of the n_TOF collaboration at CERN [12]. The application of precision printing systems like DoD inkjet printing has recently spread increasingly in many branches of the natural sciences. The DoD technique is widely used in commercial printers for producing droplets of small, well defined volumes and to dispense them in a defined area, e.g., for barcode labeling. Here, single droplet dosage is performed either piezoelectrically or thermally [13]. By using modified commercially available inkjet printers, the thermal DoD method and its produced deposits were investigated by Fittschen et al. [14]. Thermal DoD printers are able to produce droplets of solutions in the range of picoliters with a good reproducibility, which lead to deposits with diameters of 5 μm to 20 μm in a hemispherical shape. In contrast, simple droplet evaporation of larger droplets is known to produce ringshaped residues [14], which can best be overcome by depositing on superhydrophobic surfaces [8], which, though, introduces new limitations. For radiochemical applications, thermal DoD printers have a big disadvantage because a part of the solution is volatilized inside the printing head and the internal reservoir gets contaminated and cannot be
* Corresponding author at: Johannes Gutenberg-Universität Mainz, Institut für Kernchemie, Fritz Strassmann Weg 2, 55128 Mainz, Germany.
E-mail address: [email protected] (R. Haas). http://dx.doi.org/10.1016/j.nima.2017.08.027 Received 29 June 2017; Received in revised form 14 July 2017; Accepted 15 August 2017 Available online 24 August 2017 0168-9002/© 2017 Elsevier B.V. All rights reserved.
R. Haas et al.
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49
exchanged. Hence, the printing head would have to be fully exchanged for a different application to avoid cross-contamination effects. As an alternative to thermal inkjet printing piezoelectric DoD dispensers have been investigated. One such system is the PipeJet® P9 Nanodispenser from Biofluidix. This features an external reservoir, which is in contrast to thermal inkjet printers. It was shown that the modified thermal inkjet printer has a smaller deposition volume as well as the ability to print bands, whereas the piezoelectric dispenser has a slightly better precision in printing spots [15]. Piezoelectric DoD printers are currently used, e.g., in energy research to produce inkjet-printed Cu2 ZnSn(S, Se)4 solar cells [16]. Piezoelectric dispensers have an external reservoir, which is easily exchangeable and the solution is not in contact with internal parts of the dispenser. Therefore, in this work, a new method to produce targets based on piezoelectric DoD inkjet printing has been developed and basic features of the produced thin layer deposits have been characterized. The setup of the system is built up of a sample stage and a dispenser based on the DoD technology with an external reservoir for liquids. It is possible to dispense liquids over a wide range of viscosities, from water-based solutions to, e.g., glycerin-based ones. Droplet volumes between 5 nL and 60 nL can be produced, which is far below the limit of commercially available manual pipettes. In this work, different droplet volumes for water-based solutions with a specific concentration were verified using radioactive tracers. Additionally, the residue after droplet evaporation was characterized via optical microscopy as well as scanning electron microscopy (SEM). The system was finally applied to produce well defined patterns of radioactive species to study the spatial resolution of an autoradiographic imaging system as a function of 𝛽-decay energy.
Fig. 1. View of the DoD printing system containing the piezo dispenser (1) and two compact motorized translation stages (2) with a holder for substrates with a diameter of 26 mm (3).
2. DoD inkjet printing system 2.1. Setup of the printing system The inkjet printing system comprises a DoD dispenser (Biofluidix, PipeJet® P9 Nanodispenser) for the droplet production, and two compact motorized translation stages (Thorlabs) to move a substrate in two dimensions, see Fig. 1. The dispenser consists of a printing head with an electronically controlled piezo-driven piston inside its casing and a replaceable tip connected to an external reservoir. The tips are made of polyimide and polypropylene and have an inner diameter of 200 μm. To connect the tip and the reservoir, an 8 cm-long flexible tube (Proliquid GmbH, LMT-55) made from TYGON® is used. The reservoir is a commercially available pipette tip made of polypropylene and can be fixed to the holder of the dispenser. To adjust the height of the dispenser over the substrate, the former is mounted to a stand. Each translation stage has a travel range of 50 mm and a bidirectional repeatability of 1.6 μm [17]. To fix substrates in a well-defined position during movements, a holder for circular substrates with a diameter of 26 mm is mounted on the top stage. This can be easily replaced with holders for targets of different geometries, providing for the flexibility of the setup. The liquid is held in external, easily exchangeable and cheap parts and not in the casing of the dispenser, which is a significant advantage over normal inkjet cartridges and especially relevant for applications using a multitude of sample solutions. Thus the work with radioactive substances is simplified, because the contaminated parts can easily be exchanged. For good printing results, a vertical distance of 1 mm to 2 mm between the tip and the substrate is recommended by the manufacturer. Liquids and solutions with a viscosity of 0.5 mPa⋅s to 500 mPa⋅s and a surface tension of 30 mN/m to 76 mN/m are suitable for dispensing. The manufacturer specifies a volume precision of <3% and accuracy of <10% for the dispenser [18]. Drop volumes of 5 nL to 60 nL can be generated in dependence of the used tip diameter. For each drop volume, concentration of a solution and solvent, the stroke velocity of the piston has to be calibrated via a manufacturer-delivered software. An optimal value is found when the dispenser can continuously print without
Fig. 2. Schematic representation of the dispensing process [18]. A liquid is kept inside the elastic polymer tip by its surface tension until a fast displacement of the piezo driven piston. The liquid is subsequently pressed to both directions, and a droplet forms at the lower end of the tip and is released. After a slow release of the piston, new liquid runs from the reservoir into the tip to fill up the tip again.
generating satellite drops. The drop volume is set by the stroke range and the stroke velocity of the piezo-driven piston inside the dispenser. The schematic principle of the dispenser is illustrated in Fig. 2. Once filled, the liquid is kept inside the elastic polymer tip due to its surface tension. After a fast displacement of the piston with a specific stroke range and stroke velocity, the liquid is forced out of the respective part of the tip hit by the piston in both directions. At the bottom of the tip, a drop is generated and falls off. By a slow release, new liquid from the reservoir fills up the tip again. The stroke range influences the suppressed liquid volume inside the tip and hence is crucial for setting a drop volume. Additionally, the stroke velocity enables dispensing of a specific volume. If the stroke velocity is set too low, the kinetic energy is insufficient to let the formed drop fall off the tip. Too high values cause spraying or the generation of satellite drops. Therefore, the stroke velocity has to be calibrated for each solution and desired volume. Typical stroke velocity values are 60 μm∕ms to 100 μm∕ms for water-based samples, 50 μm∕ms to 80 μm∕ms for organic solvents and >150 μm∕ms for viscous media. The maximum frequency of the dispenser to produce droplets lies at 100 Hz depending on the liquid used. After the dispensing process the drops evaporate on the substrate, leaving behind the residue of the dissolved species. 44
R. Haas et al.
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49
were performed with stroke velocities of 100 μm∕ms and 145 μm∕ms. To obtain enough activity with a single test tube for gamma spectroscopy, 100 drops were dispensed in each of the test tubes. Additionally, 1 μL of the radioactive solution was dispensed into a third test tube with a pipette (Eppendorf, reference 2, 0.5 μL to 10 μL) as a standard volume. Every test tube was filled with isopropanol to 1 mL to obtain a similar sample geometry, and was measured with a gamma detector (Canberra, GX6020, rel. eff. 64.8%) for 30 min. The results obtained with 140 La were verified in a second identical series of measurements, where the 198g Au tracer was used. For this 500 μL of the 198g Au solution was filled into the reservoir of the dispenser and different volumes and velocities were dispensed under the same conditions as the 140 La tracer. To eliminate any influence of the set stroke velocity for different volumes, the stroke velocity was kept fixed at 90 μm∕ms. Every sample was measured for 10 min with the gamma detector used before. Scanning electron microscopy (Philips Inc., XL 30) was used to investigate the distribution pattern of the deposits. For this, an unirradiated solution of the La(NO3 )3 was dispensed in a rectangular pattern onto a titanium foil and a graphene foil. The settings of the dispenser were 8 nL volume with a stroke velocity of 100 μm∕ms. The sequence was a 10 × 10 pattern of points with an interval of 0.5 mm.
Fig. 3. Example of a planned sequence preview with equidistant drops in a circular area with 5 mm diameter.
2.2. Software To ensure a fully automated process of printing, including the synchronization of the movement of the stages and the timing of the dispenser, a custom software suite based on LabView has been developed. It uses a two-axis coordinate system with a grid resolution of 0.01 mm and can support up to 64000 individual drops. A series of coordinates is referred to as a sequence and represents all information needed to print a target. Sequences can be exported to files for later use and sharing. The planning of sequences can also be done independently of the printer by a standalone software ‘‘planning tool’’ to increase efficient usage of the actual printer hardware. The possibilities to use polar coordinates or to create repeatable patterns of coordinates are also integrated in the planning tool. Furthermore, the software suite provides the user with a preview of the planned sequence and is able to simulate the printing process for testing and controlling purposes. An example of a planned sequence preview consisting of equidistant drops, cut to fit a circular 5 mm target, is given in Fig. 3. A further function is the conversion of graphical coordinates into Cartesian coordinations, which allows the user to transform a monochrome bitmap into a sequence.
2.4. Results and discussion 2.4.1. Drop volume The results of the determination of the drop volume using 140 La tracer are shown in Table 1 and give a precision of (5.0 ± 2.5)% for a volume of 8 nL. The results of the 198g Au tracer verify the precision of the 140 La tracer with an overall precision of (5.2 ± 1.4)%. The influence of the stroke velocity on the drop volume is clearly visible in the data given in Table 1. The deviation of the drop volume of nearly 15% for three different stroke velocities affirms the importance of a calibration for a selected stroke velocity before printing. Otherwise, the dispensed drop volume may be up to 40% higher than the preset volume (cf. Table 1). The volume also depends on the viscosity of the used solution, which depends on the concentration of the dissolved salt and the used solvent. The average deviation of three measured samples for each drop volume is 1.5%, which confirms a good reproducibility, once the calibration is performed. This is essential to produce samples containing a well-defined amount of deposited material.
2.3. Production and characterization of DoD targets 2.4.2. Deposition dimensions on Ti foils and on graphene SEM pictures of lanthanum nitrate deposits are shown in Fig. 4. They are visible by the material contrast of the SEM pictures. Thus, the deposits appear darker than the titanium substrate (a) and brighter than the graphene substrate (b). While the deposits on titanium (a) are irregular and have diameters of about 400 μm, the ones on graphene (b) have a smoother and more regular surface and have much smaller diameters of about 170 μm. The structure of the titanium foil is visible through the deposit in (a), indicating this to be very thin. The thickness of the deposits can be determined from the concentration of the solution, the dispensed drop volume and the diameter of the deposit. The deposits on Ti foil have an average areal density of about 16 μg⋅cm-2 , and those on graphene foil about 88 μg⋅cm-2 . On graphene, they seem to have a hemispherical shape, as follows from the contrast in height within the deposition, which is seen in addition to the material contrast of the SEM image. This shape and the smaller diameter result from different surface properties. The graphene foils have a more hydrophobic surface than the untreated titanium foils [19]. This property causes a higher contact angle between droplets and the substrate, which has already been investigated for the preparation of targets on superhydrophobic surfaces [8]. By this, the generation of salt rings and other defects can be avoided during evaporation of the droplets.
2.3.1. Reagents and materials 140 La (E = 1.4 MeV, E = 1596 keV, t 198g Au 𝛽 𝛾 1∕2 = 40.272 h) and (E𝛽 = 1.0 MeV, E𝛾 = 412 keV, t1∕2 = 2.69 d) were used as radioactive tracers. A solution was prepared by dissolving lanthanum nitrate (natural isotopic composition) in 0.1 mol/L nitric acid to a concentration of 2.5 mg/mL with regard to lanthanum. The radioactive tracer 140 La was produced by a 6 h irradiation of 1 mL of the La(NO3 )3 solution with thermal neutrons (flux 0.7 × 1012 n⋅cm-2 ⋅s-1 ) in the research reactor TRIGA Mainz. 198g Au was produced by irradiation of an ICPMS standard solution (Merck) of 1 mg/mL Au in 10% HCl for 6 h with the same conditions like 140 La. Circular polycrystalline graphene foils (thickness 12 μm) used as substrates for the investigation of the deposits were purchased from Applied Nanotech Inc. and had a diameter of 26 mm. Titanium foils (thickness 50 μm) with a diameter of 25 mm were also used as substrates and were washed with isopropanol before usage. 2.3.2. Instrumentation and methods To determine the dispensed volume, a small test tube made from poly(methyl methacrylate) (PMMA) was fixed under the dispenser. The reservoir of the dispenser was filled with 500 μL of the radioactive 140 La solution. The drop volume was set to 8 nL. Two measurements 45
R. Haas et al.
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49 Table 1 Analysis of the dispensed drop volume for different tracers, set volumes and stroke velocities. Tracer
concentration (g/L)
Set drop volume (nL)
Stroke velocity (μm/ms)
Measured drop volume (nL)
140 La
2.5 2.5 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
8 8 8 8 8 10 12 15 17 20
100 145 70 90 110 90 90 90 90 90
8.4 ± 0.2 8.7 ± 0.9 8.5 ± 0.1 9.6 ± 0.3 11.4 ± 0.1 10.3 ± 0.4 12.7 ± 0.1 15.8 ± 0.1 17.5 ± 0.1 21.6 ± 0.2
140 La 198g Au 198g
Au Au 198g Au 198g Au 198g Au 198g Au 198g Au 198g
These plates consist of a layer of BaFBr:Eu2+ crystals with grain sizes of about 5 μm as well as a supportive and a protective layer [20]. By exposure to a radioactive sample, the electrons of the crystals get trapped in an empty lattice of the F or Br ion. After a scan with a 650 nm HeNe-Laser the spatial information can be read out by phosphorescence of the luminescence center Eu2+ [21]. The minimum resolution of these plates depends on the chosen readout pixel size. For198g Au (E𝛽 = 1.0 MeV, E𝛾 = 412 keV) a spatial resolution below 200 μm was found by measuring a radioactive gold wire with a thickness of 2.5 μm [22]. In the present publication the resolution was determined by a different technique. For this, patterns of tracers with different 𝛽-decay energies were printed onto graphene foils with the developed DoD printing system. This technique to acquire the spatial resolution was used last by Johnston et al. with 32 P, 14 C and 35 S on older storage phosphor plates [23]. The limits of the spatial resolution were set at the Rayleigh criterion, when the saddle point between two spots is at 8/𝜋 2 ≈ 80% of the maximum intensity, and at 50% of the maximum intensity. The spatial resolution of the actual storage phosphor plates is determined by the same principle. 3.2. Experimental 3.2.1. Reagents and materials To obtain a most homogeneous deposition with every drop, circular graphene foils (thickness 12 μm, Applied Nanotech Inc.) were used as substrates. Three different tracers were used, namely (i) the low-energy pure 𝛽 - -emitter 35 S (E𝛽 = 0.2 MeV, t1∕2 = 87.5 d), (ii) the high-energy pure 𝛽 - -emitter 32 P (E𝛽 = 1.7 MeV, t1∕2 = 14.26 d) and (iii) the 𝛽 - /𝛾emitter 198g Au (E𝛽 = 1.0 MeV, E𝛾 = 412 keV, t1∕2 = 2.69 d). To produce the 198g Au tracer an ICP-MS standard solution (Merck) of 1 mg/mL Au in 10% HCl was used. The low-energy 𝛽-tracer 35 S was produced in the 35 Cl(n, p)35 S reaction on the large amount of chlorine by irradiation of the Au/HCl solution. It was induced by the epithermal part of the TRIGA neutron spectrum. A solution of NH4 H2 PO4 in millipore water with a concentration of 50 mg/mL with regard to phosphor was prepared to produce the 32 P tracer. 1 mL of both solutions were irradiated for 6 h in the TRIGA-Mainz research reactor. To get a solution with only 35 S as tracer, the gold solution was reused after about 40 days decay time (16 half-lifes of 198g Au), with only the long-lived 35 S-activity remaining.
Fig. 4. SEM pictures of lanthanum nitrate deposits on a titanium foil in (a) and on a graphene foil in (b). The drops had a volume of 8 nL and were printed with a stroke velocity of 100 μm∕ms.
3.2.2. Instrumentation and methods Scanning electron microscopy (Philips Inc., XL 30) was used to investigate the average diameter of the printed drops. Therefore, patterns of both unirradiated solutions were printed with 10 × 10 drops in an 0.5 mm interval onto graphene foils. The dispenser was set for a drop volume of 8 nL for both solutions and a stroke velocity of 90 μm∕ms for the 35 S tracer, 80 μm∕ms for the 198g Au tracer, and 68 μm∕ms for the 32 P tracer. These settings were retained for printings with the irradiated solutions. The irradiated solutions were printed in equidistant line patterns with increasing distances between two drops (cf. Fig. 5). The pattern consisted of rows with ten drops each and a distance of 1 mm between two rows. The smallest achievable distance between two drop centers without coalescing, as determined in separate studies with the
3. Application of the DoD system: determination of the spatial resolution of an autoradiographic imaging system 3.1. Introduction The print of patterns with different spacings enables to determine the spatial resolution of a radiographic imager (Fujifilm, Typhoon™ FLA-7000) with this system. Here, various imaging plates are available, depending on the kind of radioactive source used for imaging. Our imaging plates (Fujifilm, BAS IP SR 2040) are designed for 𝛽/𝛾-emitters. 46
R. Haas et al.
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49
achieve a good contrast. After every sample exposure the imaging plates were exposed to light with an IP Eraser 3 from Fujifilm for reutilization. The intensity scale of the resulting images was converted into a colored scale and the intensity of all rows was evaluated as height profiles. 3.3. Results and discussion Radiographic images of the 35 S (a), 32 P (b), and 198g Au (c) pattern are given in Fig. 6. The color scale represents the amount of absorbed energy and is only qualitative, as for all radiographic images. It shows even fine differences in the intensity between two deposits. Along the framed rows the intensity was evaluated as a height profile. Intensity profiles of three rows of the 198g Au tracer are given in Fig. 7 as an example. In profile 2 the saddle between two deposits is at 50% of the maximum intensity, in profile 3 it is at more than 80% of the maximum intensity. Every radiographic image was evaluated by these criteria. These limits are reached in rows 9 and 12 of (a), rows 8 and 12 of (b), and rows 7 and 11 of (c). The diameters of the deposits were measured in SEM pictures of further unirradiated patterns and were about (500 ± 16) μm for the sulfur deposition, (260 ± 10) μm for phosphor, and (374 ± 10) μm for gold. Due to the distance between two drop centers, the resulting spacing between two 35 S deposits of the measured rows is (250 ± 16) μm in (a.1), (163 ± 16) μm in (a.2), and (72 ± 16) μm in (a.3). By the same evaluation, the spacings of the 32 P deposits are (490 ± 10) μm in (b.1), (315 ± 10) μm in (b.2), and (215 ± 10) μm in (b.3) and the spacings of the 198g Au deposits are (376 ± 10) μm in (c.1), (226 ± 10) μm in (c.2), and (126 ± 10) μm in (c.3). Following the principle of Johnston et al. [23] to determine the resolution limits of the plates, the Rayleigh criteria and the 50%-limits of the investigated tracers are given in Table 2. Because the individual droplets deposits are still easily identifiable by the naked
Fig. 5. The sequence used for the determination of the spatial resolution of the storage phosphor plates.
Au solution, was 400 μm. This distance was subsequently incremented by 25 μm per each row up to a maximum of 750 μm. Radiographic Imaging was performed with a Typhoon™ FLA-7000 from Fujifilm using (Fujifilm, BAS IP SR 2040) imaging plates to investigate the spatial resolution for different radiation types and energies. To avoid any contamination of an imaging plate, a Mylar® foil was set between the sample and the plate. The exposure time was chosen in relation to the activity of the measured sample. For samples with high activity, e.g., 198g Au the exposure time was set to a few minutes whereas for samples with low activity, e.g., 32 P, it was set to up to two hours to
Fig. 6. Radiographic images of the 35 S (a), 32 P (b), and 198g Au (c) pattern on the left and REM image (d) as well as light microscope images (e, f) for comparison on the right. The sample exposure time was 2 h for 32 P, 1 h for 35 S and 5 min for 198g Au. The color scale qualitatively represents the amount of absorbed energy and is shown relative to the maximum intensity. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
47
R. Haas et al.
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49 Table 2 Limits of the spatial resolution for three tracers of different 𝛽-decay energies and the value from Ref. [22]. tracer
𝛽-decay energy (MeV)
50%-limit (μm)
Rayleigh criterion (μm)
Ref. [22] (μm)
32 P
1.7 1.0 0.2
315 ± 25 226 ± 25 163 ± 25
215 ± 25 126 ± 25 72 ± 25
– <200 –
198g Au 35 S
with higher decay energy. Considering just the 𝛽 radiation, the spatial resolution for 198g Au should be between those of 32 P and 35 S, which is the case. So the impact of the gamma radiation is not significant enough to cause a substantially lower spatial resolution for the 198g Au tracer. Also the Rayleigh criterion of 198g Au with (126±25) μm fits to the spatial resolution found by Liebe et al. of < 200 μm [22]. 4. Conclusion A novel method for the preparation of thin layer targets was developed on the basis of the Drop-on-Demand technology. A printing system was designed using a piezoelectric dispenser and a x,y-displaceable sample stage. A software was developed to enable automated printing of defined patterns. The dispense precision as well as the shape of deposits on different substrates were investigated with radioactive tracers and SEM. The dispensed drop volume is in agreement with that indicated by the manufacturer for the settings of the dispenser with an overall precision of about (5.2 ± 1.4)%, but was found to depend significantly on the used stroke velocity. On graphene foils the depositions result in nearly homogeneous and hemispherically shaped residues with diameters below 500 μm at a drop volume of 8 nL. Thus the formation of salt rings can be avoided at such small drop volumes. The thickness of the deposits can be minimized by a low concentration and large drop volumes. As a first application of the printing system, the spatial resolution of a radiographic imaging system was investigated with solutions of 32 P, 198g Au and 35 S tracers. With the high accuracy of the printing system, the spatial resolution was measured in relation to the 𝛽-decay energy, which would not easily have been achievable by manual pipetting. For the lowest energy of 0.2 MeV (35 S) the limit of the spatial resolution is at (72 ± 25) μm when defined according to the Rayleigh criterion. The developed printing system establishes a new method for precise target preparations, which overcomes limits of conventional preparation methods, with the ability of automated printing and a high accuracy. Acknowledgments The authors acknowledge the local support of the staff of the TRIGA Mainz for performing the irradiations and of the mechanical workshop at the Institute of Nuclear Chemistry. We acknowledge the financial support from the Helmholtz Institute Mainz. References [1] J. Runke, Ch.E. Düllmann, K. Eberhardt, P.A. Ellison, K.E. Gregorich, S. Hofmann, E. Jäger, B. Kindler, J.V. Kratz, J. Krier, B. Lommel, C. Mokry, H. Nitsche, J.B. Roberto, K.P. Rykaczewski, M. Schädel, P. Thörle-Pospiech, N. Trautmann, A. Yakushev, Preparation of actinide targets for the synthesis of the heaviest elements, J. Radioanal. Nucl. Chem. 299 (2014) 1081–1084. [2] B.W. Filippone, M. Wahlgren, Preparation of a 7 Be target via the molecular plating method, Nucl. Instrum. Methods A 243 (1986) 41–44. [3] A. Vascon, J. Runke, N. Trautmann, B. Cremer, K. Eberhardt, Ch.E. Düllmann, Quantitative molecular plating of large-area 242 Pu targets with improved layer properties, Appl. Radiat. Isot. 95C (2014) 36–43. [4] H.L. Adair, P.R. Kuehn, Preparation of 252 Cf neutron and fission-fragment sources, Nucl. Instrum. Methods 114 (1974) 327–332. [5] K. Eberhardt, Ch.E. Düllmann, R. Haas, C. Mokry, J. Runke, P. Thörle-Pospiech, N. Trautmann, Actinide targets for fundamental research in nuclear physics, in: Physics Procedia, Physics Procedia: 28th World Conference of the International Nuclear Target Development Society, (accepted for publication), 2017. [6] W. Parker, R. Falk, Molecular plating: A method for the electrolytic formation of thin inorganic films, Nucl. Instrum. Methods 16 (1962) 355–357.
Fig. 7. Linear profiles of the intensity along rows 1, 2 and 3 of the radiographic image of 198 Au as labeled in Fig. 6. The distances of the drop centers were (376 ± 10) μm in row 1, (226 ± 10) μm in row 2, and (126 ± 10) μm in row 3.
eye at the 50%-limit in the radiographic images, it seems more suitable to use the Rayleigh criterion as the limit for the spatial resolution of the used RI system. The results point to lower spatial resolution for tracers 48
R. Haas et al.
Nuclear Inst. and Methods in Physics Research, A 874 (2017) 43–49 [14] U.E.A. Fittschen, N.H. Bings, S. Hauschild, S. Forster, A.F. Kiera, E. Karavani, A. Fromsdorf, J. Thiele, G. Falkenberg, Characteristics of picoliter droplet dried residues as standards for direct analysis techniques, Anal. Chem. 80 (2008) 1967–1977. [15] G.E. Morlock, C. Oellig, L.W. Bezuidenhout, M.J. Brett, W. Schwack, Miniaturized planar chromatography using office peripherals, Anal. Chem. 82 (2010) 2940–2946. [16] X. Lin, J. Kavalakkatt, M.C. Lux-Steiner, A. Ennaoui, Inkjet-printed Cu2 ZnSn(S, Se)4 solar cells, Adv. Sci. 2 (2015) 1500028. [17] Thorlabs GmbH, MTS Series User Guide, 2016. [18] BioFluidix GmbH. PipeJet P9 Nanodispenser Datasheet, 2014. URL www.biofluidix. com. [19] K. Balasubramanian, M. Burghard, Chemie des Graphens, Chem. Unserer Zeit 45 (2011) 240–249. [20] GE Healthcare Life Sciences. Storage Phosphor Screen BAS-IP, 2012. URL https: //www.gelifesciences.com. [21] H. von Seggern, Photostimulable x-ray storage phosphors: A review of present understanding, Braz. J. Phys. 29 (1999) 254–268. [22] D. Liebe, K. Eberhardt, W. Hartmann, T. Häger, A. Hübner, J.V. Kratz, B. Kindler, B. Lommel, P. Thörle, M. Schädel, J. Steiner, The application of neutron activation analysis, scanning electron microscope, and radiographic imaging for the characterization of electrochemically deposited layers of lanthanide and actinide elements, Nucl. Instrum. Methods A 590 (2008) 145–150. [23] R.F. Johnston, S.C. Pickett, D.L. Barker, Autoradiography using storage phosphor technology, Electrophoresis 11 (1990) 355–360.
[7] A. Vascon, S. Santi, A.A. Isse, T. Reich, J. Drebert, H. Christ, Ch.E. Düllmann, K. Eberhardt, Elucidation of constant current density molecular plating, Nucl. Instrum. Methods A 696 (2012) 180–191. [8] D. Renisch, T. Beyer, K. Blaum, M. Block, Ch.E. Düllmann, K. Eberhardt, M. Eibach, S. Nagy, D. Neidherr, W. Nörtershäuser, C. Smorra, Targets on superhydrophobic surfaces for laser ablation ion sources, Nucl. Instrum. Methods A 676 (2012) 84–89. [9] M.A. Garcia, M.N. Ali, T. Parsons-Moss, P.D. Ashby, H. Nitsche, Metal oxide films produced by polymer-assisted deposition (PAD) for nuclear science applications, Thin Solid Films 516 (2008) 6261–6265. [10] M.A. Garcia, M.N. Ali, N.N. Chang, T. Parsons-Moss, P.D. Ashby, J.M. Gates, L. Stavsetra, K.E. Gregorich, H. Nitsche, Metal oxide targets produced by the polymerassisted deposition method, Nucl. Instrum. Methods A 613 (2010) 396–400. [11] Q.X. Jia, T.M. McCleskey, A.K. Burrell, Y. Lin, G.E. Collis, H. Wang, A.D.Q. Li, S.R. Foltyn, Polymer-assisted deposition of metal-oxide films, Nat. Mater. 3 (2004) 529– 532. [12] E.A. Maugeri, S. Heinitz, R. Dressler, M. Barbagallo, N. Kivel, D. Schumann, M. Ayranov, A. Musumarra, M. Gai, N. Colonna, M. Paul, S. Halfon, L. Cosentino, P. Finocchiaro, A. Pappalardo, Preparation of 7 Be targets for nuclear astrophysics research, J. Instrum. 12 (2017) P02016. [13] B. Derby, Inkjet printing of functional and structural materials: Fluid property requirements, feature stability, and resolution, Annu. Rev. Mater. Res. 40 (2010) 395–414.
49