Flexible printed cables are everywhere in modern physics experiments that hunt for rare events, from dark matter detectors to searches for neutrinoless double beta decay. Their copper-polyimide construction offers high resistivity, high dielectric strength, flexibility, and stability across wide temperature ranges, making them ideal for the ultra-high vacuum and cryogenic environments common in low background experiments. Yet these same cables have long carried a hidden burden: radioactive contaminants at levels that can swamp the extremely faint signals such experiments are designed to detect. A new study now reports a systematic way to strip that burden away, producing fully functional cables with uranium-238 and thorium-232 contamination tens to hundreds of times lower than commercial alternatives.
Past surveys of commercial copper-polyimide flexible cables found uranium-238 concentrations of 400 to 4700 picograms per gram, thorium-232 at 16 to 3700 picograms per gram, and natural potassium at 170 to 2100 nanograms per gram. For detectors searching for neutrinoless double beta decay or direct dark matter detection, cabling is often a significant contributor to the radioactive background budget. Experiments such as nEXO and OSCURA have had to limit cabling to the absolute minimum or turn to alternative materials that, while more radiopure, lack the advantageous properties of polyimide-based cables, including ease of clean and reliable installation. Using alternatives also often means custom-made components, which increases cost, risk, and production time.
Earlier collaborations tried to tackle the problem. The EXO-200 team selected among commercial laminates, measured contamination during photolithography, investigated alternative chemicals, and worked with vendors on handling and packaging. The EDELWEISS collaboration surveyed laminates and adhesives and used a custom fabrication process. Even so, the final cables in those experiments contained uranium-238 and thorium-232 at hundreds to thousands of picograms per gram. Some of the authors of the new work had previously collaborated with DuPont to produce a custom laminate with uranium-238 below 10 picograms per gram, roughly ten times cleaner than commercial options, but fabricating cables from that laminate still introduced enough contamination to reach levels comparable to commercial cable.
In the new study, published in EPJ Techniques and Instrumentation, researchers led by Pacific Northwest National Laboratory together with Q-Flex Inc., a company specializing in custom long flexible cables, systematically investigated every step of the cable production process. Production starts from a simple laminate of copper and polyimide and then adds vias, patterns of traces and interconnections, insulating coverlays, and metallized pads for external connections. Each step exposes the material to chemicals, added copper and polyimide layers, and mechanical manipulation, and each is a potential source of radioactive contamination. The team chose a commercial Taiflex copper-polyimide laminate as the starting material because measurements showed it had radiopurity similar to the specialized DuPont material but was available in rolls, making it suitable for manufacturing long cables.
To track where contamination entered the process, the researchers produced panels of small detachable coupons alongside the actual cables. These coupons act as surrogates, containing all the relevant features such as traces, vias, and coverlays, and going through the same production steps as regular cables. Individual coupons can be detached after each step and measured for radioactive contaminants, allowing the team to assess how much contamination each step imparts. Where a spray or chemical bath was used, aliquots of the chemicals were also collected for assay. The analysis method was inductively coupled plasma mass spectrometry, or ICP-MS, which allows ultrasensitive assays with sub-picogram-per-gram sensitivities for uranium-238, thorium-232, and natural potassium in small samples under one gram, with quick turnarounds that enabled detailed systematic investigations of large numbers of samples.
The measurements revealed that despite starting with clean material at a few picograms per gram of uranium-238 and thorium-232, the final cable produced by the standard process reached roughly 7000 picograms per gram of uranium-238 and 700 picograms per gram of thorium-232, similar to previous findings from EXO-200 and EDELWEISS. Photolithography, the patterning of electrical traces on the copper surface using a photoresist, developing, etching, and stripping sequence, emerged as a major source, with thousands of picograms per gram of uranium-238 measured after stripping. The chemical solutions used in those steps, including anhydrous sodium carbonate, hydrochloric acid, and a potassium hydroxide based solution, were found to contain significant contamination themselves. Rather than purify or replace those carefully balanced chemicals, which would have required significant research and development and been impractically expensive, the team developed a post-cleaning method.
That cleaning recipe proved surprisingly effective. Because the lithographic process is almost entirely subtractive, removing copper from the outer layers without adding material, contaminant species from processing should sit at or near the surface. After testing various diluted acid solutions, the researchers turned to an alkaline treatment informed by literature showing that alkaline solutions modify the chemical structure of polyimide surfaces. The final procedure submerges samples in 5 percent tetramethylammonium hydroxide for 15 minutes, rinses in deionized water, dips in 2 percent nitric acid for about a minute, rinses again, and air dries in a class 10 laminar flow hood, with a soft nylon brush used for scrubbing. Repeating the entire cycle three times reduced uranium-238 and thorium-232 in a test coupon to roughly 20 and 10 picograms per gram respectively, and optical inspection at up to 500 times magnification showed no visible structural changes to the copper traces.
Other contaminating steps required different fixes. Copper plating of vias, the holes that connect conductive layers, normally uses an electroless seeding process that deposits a conductive layer over the entire cable surface, trapping contamination where cleaning cannot reach it. The team switched to a newer shadow plating process in which a thin carbon layer adheres only to the polyimide, not the copper, using fewer chemicals and depositing less material over a smaller area. Coupons made this way came out roughly nine times cleaner in uranium-238 than the electroless approach, with contamination removable by the cleaning method. Sanding with abrasive pads was found to increase thorium-232, presumably from abrasive material implanted into the surface, so the team switched to silicon carbide pads known to be generally radiopure, achieving roughly a tenfold reduction in thorium-232 at that step. For coverlays, the insulating layers applied over outer surfaces, commercial options varied by orders of magnitude, and acrylic adhesive based coverlays proved cleanest; the MCL coverlay chosen had uranium-238 and thorium-232 lower by factors of roughly 200 and 30 compared to the Taiflex coverlay.
Combining all these modifications, the researchers produced two sets of fully functional cables representing two ends of the design spectrum. The first, a simplified single-sided design developed for silicon photomultiplier readout in the nEXO neutrinoless double beta decay experiment, has traces on one side and a continuous ground plane on the other, with no vias, coverlay, or surface metallization to keep contamination minimal. The second, developed for charge-coupled device readout in the DAMIC-M dark matter experiment, includes multiple traces, copper-plated vias, coverlays on both sides, and electroless nickel immersion gold metallization on the pads. The final silicon photomultiplier cable measured roughly 20 picograms per gram of uranium-238, thorium-232 below the sensitivity of about 10 picograms per gram, and natural potassium at roughly 36 nanograms per gram. The more complex charge-coupled device cable came in at roughly 31 and 12 picograms per gram of uranium-238 and thorium-232 respectively, with higher potassium consistent with the intrinsic contamination of the MCL coverlay used.
As a direct comparison, commercial cables of the same designs showed uranium-238 and thorium-232 levels roughly 100 times and 10 times larger respectively than the custom low radioactivity versions. Compared with cables used in previous rare event experiments by EXO-200, EDELWEISS, and DAMIC, the new cables are at least an order of magnitude better in uranium-238 and thorium-232, and lower in natural potassium as well. The researchers note that all subsample replicates in the custom final cables were consistent with each other, indicating the cleaning procedure likely removed non-uniform surface contamination. They believe these levels are sufficiently radiopure for planned upgrades of currently running experiments such as DAMIC-M and SuperCDMS and future proposed experiments such as nEXO and Oscura, and that low radioactivity cables can increase experimental sensitivity not only by reducing the background budget but also by allowing additional sensors and simpler readout designs within the same budget. The same coupon-based investigative technique, they suggest, could serve as a template for investigating radiocontaminants in other material production processes, including future work on low radioactivity superconducting cables and printed circuit boards.
Subject of Research: Development of ultra-low radioactivity copper-polyimide flexible printed cables for rare-event physics experiments
Article Title: Ultra-low radioactivity flexible printed cables
Article References: Arnquist, I. J., di Vacri, M. L., Rocco, N., Saldanha, R., Schlieder, T., Patel, R., Patil, J., Perez, M., & Uka, H. (2023). Ultra-low radioactivity flexible printed cables. EPJ Techniques and Instrumentation, 10(1), Article 17. https://doi.org/10.1140/epjti/s40485-023-00104-6
Image Credits: AI Generated
DOI: 10.1140/epjti/s40485-023-00104-6
Keywords: radioactivity, flexible printed cables, copper-polyimide, ultra-low background detectors, ICP-MS, photolithography, shadow plating, coverlay, dark matter, neutrinoless double beta decay, radiopurity, materials science
News Source: Denise Maddox. (October 30, 2026). Physicists engineer flexible cables clean enough for next-generation dark matter detectors. Scienmag.
