Fig. 1. Schematic diagram of the Borexino detector in cross-section.

Fig. 2. Prototypes of the nested nylon vessels. They were built in a fashion similar to that of the real vessels, although not under clean conditions. The test inflation shown here was done in the Jadwin Gymnasium at Princeton University.

Fig. 3. Schematic diagrams (not to scale) of stress–strain relationships for nylon films below the glass transition temperature (left) and above it (right). Typical values for the tensile strength σ_t are 75 MPa for film in the glassy state, and 20 MPa for film in the plastic state.

Fig. 4. The accordion-style stack technique used for constructing the nylon vessel envelopes. The edges of wedge-shaped nylon panels are bonded together to form the spherical vessel; the bond is temporarily clamped down against the working surface to cure. Panels are folded into a stack once bonded. The process continues until all panels of a vessel are in the stack. (In addition to permitting vessel construction within the limited confines of the clean room, as a bonus the stacking permits the film to be self-covering, reducing exposure to radon daughters in the air.) The last joint is conceptually identical but requires unfolding some panels in the stack to align the top and bottom panels for solvent bonding.

Fig. 5. Making a joint for the nylon panels. A first cart (left) carries the spray gun with which the Resorcinol solution is applied and a second cart (right) follows, folding the edge of the panel beneath onto the one above. The clamps are then closed to apply the necessary pressure. Close-ups of the spray and folding carts are shown.

Fig. 6. Sketch of the IV upper end region (i.e., polar) assembly in cross-section. The bonding technique for transitioning from nylon film to bulk nylon is shown in Fig. 7.

Fig. 7. Bonding scheme for the end region transition between the vessel nylon membrane and the polar assemblies. A layer of half millimeter-thick nylon film is placed on the bulk nylon surfaces with pure formic acid (dashed line). The vessel membrane can then be joined to the assembly relying only on well-tested resorcinol bonding (dotted line). The drawing refers to the IV collar ring region, but this technique is applied to all the film-to-bulk nylon transitions. The L-ring is simply bolted in placed (nylon bolts not shown for simplicity).

Fig. 8. Leak checking the IV prior to shipping. Top: mixing of SF₆ gas that was injected directly in the clean room, and the decay time of its concentration, due to dilution with new air entering the clean room, following an injected spike. The plot shows that the mixing time in the clean room is much faster than the dilution time. Bottom: the step in SF₆ concentration that occurs after an accumulation bag surrounding the SF₆-filled nylon vessel is flushed into the clean room. The size of this step after a 10 h accumulation in the bag implies an equivalent leak rate of for the IV.

Fig. 9. Deformations and stresses of the inner nylon vessel in conditions of a buoyant load at 0.5% density difference between scintillator and buffer fluids. The difference between models using elastic vs. viscoelastic analyses are shown under the two conditions of 20% and 100% relative humidities in the ambient environment.

Fig. 10. Minimum pressure required, during the gas inflation, to compensate the under-dimensioning of the rope and avoid wrinkles.

Fig. 11. The set of nylon vessels packaged in the frame for shipping to LNGS.

Fig. 12. The nylon vessels partway through the installation.

Fig. 13. The nylon vessels in the SSS after inflation with synthetic air.

Fig. 14. Graphs of the O₂ concentration in the IV during the first two stages of purging. Time in hours is shown on the horizontal, and the percentage of O₂ on the vertical. After the end of the first purging stage, the stratified gas continued to mix, so the concentration is lower at the beginning of the second stage.

Fig. 15. Radial distribution of the reconstructed positions of external γ-ray and neutrino scintillation events in the energy window 0.25–0.8 MeV. Note that signal-to-noise remains greater than one in all radial shells of scintillator only for a fiducial volume of radius or so, corresponding to a scintillator mass of . The vertical bar at 4.25 m is the position of the inner nylon vessel. Since this simulation takes the finite resolution of position reconstruction into account, some events have a reconstructed position outside the inner vessel.

Measured ²²⁶Ra in the two nylon film candidates [34]

Although the Capron sample extruded for this measurement was only thick, both Borexino vessels are made from film.

Measurements or upper limits of radioactive contaminants in the end region components that contribute most to the background. Radioactive contaminant levels are reported in parts per trillion (ppt) by mass for U and Th, parts per million (ppm) by mass for potassium (all isotopes), and as activities in mBq per kg for ⁶⁰Co

As most of these measurements were performed by early 2001, and the half-life of ⁶⁰Co is 5.3 years, we expect that the current ⁶⁰Co contamination levels are reduced by more than a factor of two.

Radioactive contaminants of various ropes

The rope selected for Borexino was the Tensylon.

Radioactive contaminants in the instrumentation

The tabulated mass of 1.1 kg for the load cells refers to only the 16 that are positioned near the IV.

Fixed-volume vessel filled to 50 Pa prior to buoyancy onset, no wrinkling, with and without creep

Using creep compliance measured for 5 MPa at 20% relative humidity, and C38F data at 7 MPa from 1992 [19].

Relative molar solubilities (volume of gas dissolved per unit volume of liquid) of Ar, Kr and Rn in water and pseudocumene at

By definition, the solubility of each in N₂ gas is 1. For example, in equilibrium the molar concentration of Ar in N₂ gas is 1/0.039=26 times its concentration in water. Data taken from values in the Solubility Data Series, extrapolated to 15 [43] and [44]. The estimated values for pseudocumene are averages of available data for the similar compounds benzene and toluene (which differ by 20%).

Gamma background in Borexino from the various components of the end region

The background is expressed as rate in events/day in the 250–800 keV energy window (NW) and in the 800–1300 keV energy window (pep). Rates are given for the entire volume of scintillator (“all”), and for 200- and 100-ton central fiducial volumes.

Measured diffusion properties of ²²²Rn in nylon as functions of relative humidity (RH)

Data from Refs. [19] and [35]; see text for details. The top two lines are for nylon film between two volumes of dry pseudocumene (PC) and for nylon separating volumes of pseudocumene and water. The diffusion length ℓ is defined in Eq. (3); the effective permeability P_eff is defined in Eq. (4) and computed for the thickness of the Borexino IV, .