The preparation of isotopic boron targets – searching for a more consistent approach

. The reliable availability of isotopic boron targets has gained in importance concerning light ion reactions withing the realm of nuclear astrophysics research and remain somewhat elusive. An exhaustive approach was undertaken via electron beam evaporation to produce self-supporting and backed targets spanning quite a few isotopic samples across several suppliers, resulting only in limited success. Details of the extensive sample preparation procedures undertaken, and the experimental deposition techniques explored will be presented.


Introduction and Motivation
For ongoing experiments in low-energy nuclear physics at the Argonne National Laboratory (ANL) Argonne Tandem Linear Accelerator System (ATLAS) Facility and especially studies involving nuclear astrophysics, isotopic boron targets are becoming more in demand. While current preparation techniques involve Physical Vapor Deposition (PVD) employing electron beam heating [1], reliable and consistent production of targets has been difficult to achieve. Here, the approach has been to research and document a standard procedure capable of consistently obtaining the targets needed for meeting the experimental demands.
There are many references on research pertaining to preparing boron films via the process of electron beam heating, beginning with Eschbach in 1972 [2] as well as others [3,4,5,6]. The results presented here extend the earlier work of Thomas [1,7]. Various alternate methods also appear in the literature, including electrodeposition by Verdingh [8] and Pauwels [9], focused ion beam sputtering by Baumann and Wirth [10], and Maier [11] and Muggleton [12]. Xu and Wang [13] employed the method of centrifugal precipitation, and Lozowski and Hudson [14] prepared thick targets via powder pressing. A new technique, High Energy Vibrational Powder Plating (HIVIPP), developed by Sugai [15] was used to produce boron films and later by Lipski, et al [16]. It should be noted that early on, boron was being considered as an alternative to carbon for use as stripper foils by many researchers, including Ramsay [17], and with Zeisler and Jaggi [18], incorporating multi-layers [19] and via laser ablation deposition [20].
Finally, a lot of effort has gone into the characterization and assay of boron film layers, for instance, as neutron dosimeters [21], at the Institute for Reference Materials and Standards as reference

Isotopic Starting Material
Boron is found in nature as compounds, primarily borax, the element being discovered in 1808 by Sir Humphry Davy. High-purity crystalline boron is obtained from the vapor phase reduction of boron trichloride with hydrogen on electrically heated filaments or via exchange distillation reactions [27] using a boron trifluoride compound (U.S. Patent 3,050,367). Its melting point is 2300 °C, but it sublimes at 2550 °C. Boron exists as two isotopes, 10 B 19.78% abundance and 11 B with 80.22 % abundance. The isotopes are available as crystalline form in high purity from Oak Ridge National Laboratory (Oak Ridge, TN, USA) and also from Eagle-Picher (EP Boron Quapaw, Oklahoma).
There exist at our disposal several legacy batch samples of various enrichments for both isotopes (age unknown) for use in preparing accelerator targets for ATLAS. Detailed assay and chemical analysis obtained for one of the ORNL enriched 10 B samples were given to be 98.4% total boron with 1.20% carbon and 0.28% oxygen along with other listed ppm impurities, the largest being H2O, Fe, Si and S. A similar analysis provided for one of the 11 10 B powder (97%) obtained from Quapaw, OK using a variety of techniques including FTIR, SEM, XRD, and TOF-SIMS and compared it to natural boron [28]. The results for the enriched 10 B powder revealed 2% mass content of boric acid as well as other minor constituents. Anecdotal evidence as well as experience in the preparation of accelerator targets has shown that, at times, deposition of natural material behaves in a fashion completely different from that of the enriched isotope. This is particularly true for the case of boron both in the production yield (i.e., non-adherence to the substrate as well as marginal success in floating foils) and ultimate thicknesses achieved. These difficulties in obtaining self-supporting foils from the isotope may possibly arise from the chemistry pathways employed during the enrichment process and thus differ from the naturally occurring starting material. A further example is the fact that crystalline boron isotope upon heating exhibits a tendency to eject from the electron beam hearth -due perhaps to volatile compound decomposition contained in the solid isotope [1]. It has been reported by Nandi et al. [29] that amorphous, and to a lesser extent, crystalline boron powder tends to absorb water vapor over time. This was empirically determined from various aged samples and proposed vacuum flushing with an inert gas (Ar) to remove this trapped H2O content. De Luca et.al. [28] proposed drying the boron material in a vacuum oven and purged with N2 at 130 °C. IR spectra obtained for amorphous nat B and crystalline 10 B before and after heating were consistent with results from Nandi et al. with the crystalline material showing less boric acid than the amorphous powder.

Experimental Procedures
As per the earlier work of Thomas at ANL [1], the powdered starting material is first pre-heated to drive out any volatile material. This heating was carried out withing the Intlvac Evaporator System [30] using a Ta crucible heated resistively from a crucible oven source. The powder was first pressed into a 0.95 cm diameter pellet using a hand press. The source is brought to approximately 1300 °C, measured using an optical pyrometer and heated until outgassing ceases (as registered by the nominal system pressure; base pressure ~ 10 -7 Torr), and stopped when deposition commences. The pellet was left to cool. It had been weighed prior to the heating and then re-weighed after removal afterwards. Table 1 provides results of several such heat-treated samples. Most of the powdered samples lost weight on the order of 15%.

Sample Preparation
Another aspect of standardization of the boron starting material is particle size. Using a calibrated set of sieves, samples were prepared with a particle size of 60 mesh (upper limit). This technique works fine for powdered samples, in fact, the nat B powder described above was found to be finer (80 mesh). However, most of the isotopically enriched materials (both 10 B and 11 B) were crystalline form. To prepare uniform powder starting samples from this crystalline form, the large crystalline chunks were first crushed using a percussion mortar and pestle. This material was further reduced in particle size using a SPEX Sample Prep ball mill and Ta ball media. For all samples, the resulting powder sample is then passed through a 60-mesh sieve, pressed into a pellet and pre-heated to remove volatile impurities.
To avoid loss of sample due to evaporation during vacuum heating, an alternate technique of pre-heating the isotopic powder samples was carried out in the target laboratory tube furnace under flowing Ar gas (1-2 cc/min). As before, the isotope powder was compacted into a pellet and heated to 1000 °C for 1 hour in a Ta crucible. The samples were weighed before and after. The results are given below for 10,11 B in Table 2. Due to the abundance of isotopic powder, an additional two 10 B samples #4700117 (ORNL) were also heated by this method and showed significantly higher weight loss after pre-heating than when heated under vacuum.

Substrate Preparation
The substrates chosen for this work are standard laboratory microscope slides, unfrosted, with dimensions 2.5 cm x 7.5 cm, obtained from Fisher Scientific Inc. (Waltham, MA, USA). Approximately 50-60 μg/cm 2 of NaCl are evaporated onto the slides similarly from a Ta crucible in a resistive heating crucible oven source, also contained within the Angstrom Deposition Tool [31] (see Fig. 1). From experience, sodium chloride was chosen over boron oxide as suggested by Ramsay [32]. For this substrate layer, four clean glass slides are placed above the evaporation source at 11 cm. The quartz crystal deposition monitor was positioned at 17 cm away, centered directly above the Ta crucible source.

Electron Beam Source
The boron evaporations (natural and isotopic) were carried out via physical vapor deposition (PVD) within the Angstrom Deposition Tool using the modified Telemark electron beam source (see, for example [33]). The natural boron was employed to carry out an initial test evaporation using deposition parameters previously determined for boron over many years at the ANL target laboratory. In a quick fashion, several self-support targets of thickness 100 µg/cm2 were easily obtained from this single evaporation, thus reaffirming the proper deposition data. No further deposition attempts with nat B were necessary. Four previously prepared NaCl coated glass substrates were placed at 12 cm distance above the source and are direct heated from a halogen quartz lamp placed above. The substrates are held at a temperature of 176 °C as indicated from the previous work by Thomas and recommended by Gursky [34]. The indicated temperature is monitored via a thermocouple placed on the back of the substrate slides and the evaporator system maintains temperature by varying the output of the lamp. The boron sample is placed in one of the four available Cu hearths of the electron beam source, having first been recompacted into a fresh pellet via hydraulically pressing under vacuum to reduce outgassing during heating.
For these depositions an arbitrary endpoint of 100 µg/cm 2 target thickness was chosen to begin as these would provide usable targets. Thicker depositions could be performed later once the standardized parameters for this work had become established. Once under vacuum and the substrate heated, the electron beam power is slowly increased. With a beam voltage of 8-9 kV a deposition rate of 0.2 kÅ/s was slowly achieved and maintained throughout as measured by the quartz crystal monitor placed to the side at a fixed 22 cm distance from the source. Deposition was stopped when the 100 µg/cm 2 endpoint was reached and the system allowed to cool, usually overnight, before the slides were extracted.

Carbon Backing Foils
Carbon backings of 20, 30, and 40 μg/cm 2 were prepared on the various target frames anticipated as being needed. They were made using the standard floating technique from carbon slides obtained from Arizona Carbon Foil Company (www.acf-metals.com) as they have proven over time to be a consistently reliable source of highquality carbon foils for use in our laboratory. For these evaporations, the deposition parameters were kept like those used for the self-support boron films.

Results and Current Progress
The thicknesses and number of successfully produced isotopic targets are presented below in Table 3. Compared to the relative ease of producing selfsupporting foils from the natural boron, these enriched targets were only obtained after several attempts using many substrates. Commonly, the failure was due to stresses in the foils during evaporation; the foils would curl and flake apart, likely due to incomplete bonding between the layers. Due to the limited success with producing self-supporting 10,11 B targets, further deposition attempts were then continued using carbon backing foils (see Fig. 2; Table 4).
Certainly, disappointing was the fact that even after pre-heating and particle size uniformity, the isotopic target production difficulties were still present compared to natural boron. The limited success outlined here means there is still progress to be sought out. The possibility of heating the powder to higher temperature for a longer duration and analytically monitoring for boric acid, may provide an avenue for improved deposition behaviour.

Conclusion
In conclusion, although multiple preparation techniques were explored for many samples of both isotopes, no direct correlation was ascertained as to successful target preparation. Pre-heating, particle size and powder grinding the boron samples provided a good approach, but the qualitative difficulties experienced with isotopic target preparation remain as compared to natural boron. It still appears that the chemical processing differences existing for the isotopic samples make producing selfsupporting films difficult to achieve. It was desired that among the many isotopic samples explored, procedures could be developed to alter their properties for producing thin target foils. There will likely be a steady demand for these targets once reliable preparation techniques can be established. Further investigations are therefore warranted.