The story begins with an unexpected triumph. During Expedition 399, which ran from April to June 2023, the team drilled a new borehole called Hole U1601C into the Atlantis Massif, an oceanic core complex located just 800 meters north of the famous Lost City hydrothermal field in the Atlantic Ocean. Previous drilling into mantle rocks had never penetrated much beyond roughly 200 meters below the seafloor, but Hole U1601C reached an astonishing 1268 meters. That surprising depth opened a rare scientific window: the chance to document how life is distributed through more than 1.2 kilometers of uplifted mantle and lower crustal rocks, spanning conditions that may range from highly favorable for microbes to potentially too extreme to support them. The borehole may even have reached zones where water-rock reactions create what researchers call a chemical kitchen, where organic compounds can be synthesized without any involvement of biology, making the detection of low-molecular-weight organic compounds a major expedition objective.

With the borehole outperforming all expectations, the scientific party found itself with far more microbiological material than anticipated: ultimately 191 whole-round core samples, each roughly 10 to 15 centimeters long, were collected from 172 cores. That abundance became an opportunity to stress-test and optimize every step of the sample-handling process. The central design principle, according to the team, was to maximize the potential for interdisciplinary collaboration, so that every gram of precious rock could serve as many scientific purposes as possible. That meant preserving geological context, enabling time-sensitive experiments, distributing representative subsamples of the same material to many research groups, and detecting any contamination introduced by the drilling process itself, all while working with sterile, organic-carbon-free techniques in a controlled environment.

The workflow begins on the catwalk, within minutes of a core reaching the deck. Cores are shaken out into split liners that have been cleaned and sterilized with isopropanol, and everyone examining them wears masks and nitrile gloves, with tools rinsed in methanol. Even jewelry is removed from hands and wrists, a striking precaution designed to prevent contamination of the rock with platinum group elements and gold that could confound later geochemical measurements. A dedicated selection team, consisting of one microbiologist, one geologist, a co-chief scientist, and the core curator, then chooses the microbiological sample, known as an MBIO sample, typically one per five-meter core. The choice is a delicate balancing act: the sample must be biologically promising, often targeting fractured zones with signs of fluid flow and porosity, while avoiding unique geological features whose permanent removal would leave a damaging gap in the core record. Deliberations between the microbiologist and geologist on the team proved vital for weighing these competing values.

Once selected, each MBIO sample is photographed in place, wrapped in acid-washed polytetrafluoroethylene sheeting, chosen because tests showed it contributes almost no lipids or amino acids that could muddy organic geochemical analyses, and rushed to the ship’s microbiology laboratory. There it is rinsed with a sterile 3.5 percent sodium chloride solution, labeled with its orientation, and photographed again on a rotating turntable that captures an image every 15 degrees, documenting the sample’s appearance before processing and helping researchers later reconstruct its relationship to the rest of the core. The sample then moves into a stainless steel rock box inside a chemical fume hood equipped with a KOACH clean air system and an electrostatic air eliminator. The fume hood’s negative pressure was dictated by administrative concerns about chrysotile fibers in the serpentinite cores rather than by ideal contamination control, and the authors note that a positive-pressure workspace would be preferable in the future.

The heart of the method is a two-box decontamination strategy. Using a hammer and chisel, the entire exterior of each whole-round sample, sometimes more than a centimeter thick, is chiseled away, and these exterior pieces are returned to the core liner so the rest of the scientific party can still describe and sample them. The revealed interior rock is transferred to a second, clean rock box and crushed along grain boundaries into millimeter-sized fragments, producing a homogenized material from which replicate subsamples are distributed for DNA sequencing, cell counts, cultivation experiments, lipid analysis, and organic chemistry. The team deliberately avoided electric tools such as Dremels, which generate dust and hydrocarbon contamination, and rejected a planetary ball mill because no practical cleaning method between samples could be devised. Even the choice of sanitizing agent was carefully reasoned: methanol, being more volatile than ethanol, is easier to remove from surfaces and less likely to interfere with downstream organic chemical measurements, and tools were not flamed because combustion of organic matter creates a complex smear of carbon compounds that is hard to track as contamination markers.

Verifying that all this effort actually worked required an ingenious contamination tracer. Instead of fluorescent microspheres, which are difficult to deliver consistently and prone to false negatives at temperatures above about 95 degrees Celsius, the team injected a perfluorocarbon tracer, perfluorodecalin, directly into the drilling fluid at a target concentration of one milligram per liter, using a high-pressure liquid chromatography pump whose injection rate was automatically tuned to the mud pump rate by software written by the ship’s technical staff. Loose rubble from each recovered core confirmed the tracer was reaching the borehole, with levels typically between 1 and 25 parts per billion. The critical test came with the MBIO samples themselves: exterior shavings frequently showed trace to 5 parts per billion of the tracer, but the homogenized interiors of nearly all samples contained zero or only trace levels. Just two interior samples had quantifiable PFT, and those were extremely low, at 0.007 and 0.028 parts per billion. Based on detection limits established during earlier expeditions, minimally contaminated samples should carry no more than about four cells derived from drilling fluid per gram of rock, a remarkable degree of cleanliness for material pulled from a kilometer-deep borehole.

Not every sample could be chiseled clean. A subset of cores from Hole U1601C was so crumbly and unconsolidated, often described as mushy, that separating exterior from interior proved impossible. Paradoxically, these porous, heavily altered rocks were among the most biologically interesting samples of the expedition, precisely because their open textures suggest extensive fluid circulation. Rather than discarding them, the team rinsed and homogenized them whole, and as expected, their tracer levels resembled those of exterior shavings. That contrast between chiseled and unchiseled samples actually strengthens the overall conclusion: it demonstrates both that the tracer detection works reliably and that manually chiseling the exterior is genuinely effective at removing drilling-related contamination. Results from the unchiseled samples will need cautious interpretation, and future DNA sequencing may help distinguish native taxa from likely contaminants, but the authors argue such samples should not be ignored.

Beyond the immediate results, the paper distills hard-won lessons for future expeditions. The team recommends tracing drilling-induced contamination in real time so shipboard procedures can be adjusted on the fly; distributing homogenized samples across the entire scientific party, not just microbiologists, so geochemical and microbiological data are generated from the very same material; staffing at least two and preferably four or more microbiologists aboard to enable rapid processing, which during Expedition 399 meant subsamples reached their final storage destinations roughly two hours after core recovery; and relying on simple tools that can be thoroughly cleaned between samples. They also flag a key gap: the lack of a true shipboard clean room. Dust particles from ambient laboratory air have been shown to be the dominant contamination source in low-biomass subseafloor rocks collected during a previous Atlantis Massif expedition, and an enclosed, positive-pressure workspace with room for multiple workers would greatly reduce that risk. With hundreds of homogenized subsamples now archived and a proven protocol on record, the legacy of Expedition 399 may shape the exploration of the deep rocky biosphere for decades to come, even as the drilling platform that made it possible sails into retirement.

Subject of Research: Sampling and contamination control methods for microbiological and biogeochemical analysis of hard rock cores from the oceanic crust

Article Title: An interdisciplinary approach to sampling hard rock cores of the oceanic crust for microbiological and biogeochemical research

Article References: Brazelton, W. J., Cavazos, O., Robare, J. A., Southam, G., Suhonen, J., Wang, F., Lang, S. Q., McCaig, A., Blum, P., Abe, N., Coltat, R., Deans, J. R., Dickerson, K. L., Godard, M., John, B. E., Klein, F., Kuehn, R., Lin, K.-Y., Lissenberg, C. J., … Wheat, C. G. (2026). An interdisciplinary approach to sampling hard rock cores of the oceanic crust for microbiological and biogeochemical research. Scientific Drilling, 35(2), 159-169. https://doi.org/10.5194/sd-35-159-2026

Image Credits: AI Generated

DOI: 10.5194/sd-35-159-2026

Keywords: oceanic crust, subseafloor biosphere, IODP Expedition 399, Atlantis Massif, Lost City hydrothermal field, microbiology, biogeochemistry, scientific drilling, contamination tracer, perfluorocarbon tracer, serpentinization, JOIDES Resolution