“Dark Oxygen”  – A Reason to Accelerate Nodule Extraction?

Andrew Sweetman and his team of researchers from Heriot-Watt University published an article on July 22, 2024, in Nature Geoscience, citing evidence that polymetallic nodules produce oxygen – what Sweetman terms “dark oxygen” – through electrolysis on the seafloor.  Sweetman cites the fact that nodules possess high enough voltage potentials to explain this hypothesized electrolysis and hence oxygen production. 

If Sweetman is correct, then these nodules serve as solid-state perpetual energy machines – unconnected to external power sources and able to power electrolysis (continuously or not) over extremely long periods of time.  The implication is that these nodules will prove even more valuable than people ever imagined.  Extracted nodules could provide a source of virtually unlimited, clean energy to society forever. 

Alas, we fear that like every other perpetual energy machine that’s ever been sold to society, this one will fall short.  Nonetheless, nodules remain a very valuable resource that can be extracted with dramatically lower impacts than alternative sources. 

Even though Sweetman’s results would advance the cause of nodule extraction to the point that just about every person on the planet would have to endorse their extraction, we remain skeptical due to some inconsistencies and irregularities in the results and work.  We could be wrong in our examination. If Sweetman’s results are reproduced by multiple independent scientists following proper protocols, we will stand corrected.  Science, however, demands skepticism.

Our concerns at a high level are as follows: 

• At least three other studies have produced results that directly contradict Sweetman’s
• Electrochemists we’ve consulted don’t believe the results are possible physically
• Sweetman’s methods have been questioned and may have led to flawed results

Additionally, at the end of this note we walk through the scenario where Sweetman’s results are correct, and nodules do indeed produce oxygen.  We put that production in perspective.  We find that the amounts of production are very small relative to the amount of oxygen produced (and CO2 captured) by ecosystems where we currently extract energy minerals today.  So, any suggestion that we should avoid nodule extraction because of this hypothesized electrolysis appears to be misplaced. 

We note that the media and environmental groups have eagerly run with this story to back the narrative that there are too many unknowns to extract nodules from the seafloor, and that the purported oxygen creation is another reason to disqualify nodule extraction.  In their zeal to stop all development everywhere at all costs, these groups seem unable to detect the perpetual clean energy promise of the nodules (written with a measure of sarcasm). 

Below we examine the science. 

What Does the Scientific Method Say?

One of the key tenets of the scientific method holds that a hypothesis is only accepted by the scientific community once it has been confirmed.  Science is designed to be skeptical.  If an experiment can’t be repeated to produce the same results, then the hypothesis may need to be altered, or discarded, or more experiments may need to be run. 

We know that Sweetman’s hypothesis is rejected by three separate studies:

• First is a 2006 study published by Alexis Khripounoff in the American Society of Limnology and Oceanography, Inc.  That study, which used the same type of respirometer used by Sweetman found that benthic samples consumed O2, in other words there was oxygen uptake as expected.  “The oxygen uptake of 0.74 mmol m^-2 d^-1 is equal to 7.1 mg m^-2 d^-1 of organic carbon mineralized by the biological activity (assuming a respiratory coefficient of 0.85).”
• Another study, cited by TMC in its press release on the Sweetman publication, was run by Sung-Uk An, and others, and published in Deep Sea Research Part I in May of 2024.  The experiment included tests in polymetallic nodule fields in the Clarion Clipperton Zone as well as two seamounts.  All samples indicated oxygen uptake, though at different rates, in direct contradiction to the Sweetman results. 
• Finally, TMC also references a study done by researchers from the University of Southern Denmark on polymetallic nodules located in the CCZ, near where Sweetman’s samples originated, that used a different type of measurement device – an Eddy Covariance lander – but that also found oxygen consumption in the nodule fields rather than oxygen production.  Based on discussions with scientists, we believe data from this study will be part of TMC’s environmental impact assessment which has not yet been released. 

Above chart from 2006 Khripounoff study

None of this is to say that Sweetman’s results won’t be confirmed in other independent tests in the future.  But given the conflicting data coming from different studies we do think it is appropriate to wait to draw conclusions.

What Do Principles of Electrochemistry Say?

We queried a few electrochemists and other ocean technology experts to get their reaction to the paper as we had questions about how the process of electrolysis could be powered for millions of years without an external power supply and without rapid growth in the nodules.  We also had questions about the ability of nodules to power electrolysis when the minimum required voltage is known to be 1.23 Volts and the nodules Sweetman studied generally held voltage potentials in the 0.05-0.15 V range with measurements as high as 0.95 Volts. 

The electrochemists with whom we consulted had various comments, but in general they were quite skeptical of the results. 

• They noted that without an external power source, it is difficult to understand how the electrolysis would work.  Electrolysis requires power and while voltage potentials could provide some power, those potentials would be dissipated as the electrolysis was performed.  The “battery” needs to be recharged or the differentials disappear.  Nodules don’t grow fast enough to compensate for the discharge. 
• More importantly, there needs to be a current, not just voltage.  One electrochemist noted that it was interesting that Sweetman didn’t measure current in his study.  That current implies that electrons/ions can flow between the cathode and the anode to form a circuit.  Voltage differentials exist everywhere in the world but if they aren’t connected in a manner that allows this flow, then the voltage differences mean nothing.  Without that requirement, just about everything in the world would function as a battery. 
• Jon O. Hellevang, Head of R&D at GCE Ocean Technology, noted that if the purported effect were true, and it applied to manganese crusts as suggested by Sweetman, that we should see an “enormous voltage level for crusts” but none has been recorded. 
• There doesn’t seem to be enough voltage differential to power electrolysis.  Hellevang also noted this in his post.  The electrochemists weren’t certain about this because it’s difficult to know exactly how voltage is being measured, or if the nodules might act together in a series, but we note that the electrolysis of water requires 1.23 Volts (Link) (with electrochemists telling us the actual limit is probably close to 2 V because of some technical factors) and Sweetman cites a max Volt measurement of 0.95 with most nodules in the 0.05 to 0.15 V range. 

Methods Questioned and Curious Results

There are some circumstances around the Sweetman study which raise yellow flags.  TMC mentions some of these in its release, and we have gathered some other from scientists who follow the research closely.  (TMC hired Sweetman, and their people would have worked closely with the scientist, so they are in a fairly good position to assess his methods.  Also, TMC’s self-interest would dictate strong support for Sweetman’s results but that support is not present, and this fact is notable).

• TMC notes that the data Sweetman used were not collected under conditions representative of the abyssal plains seafloor.  If you read through the Methods section of the Sweetman report, you’ll see that the ex-situ work was not done at pressures consistent with the abyssal plains.  TMC notes that core samples were contaminated with water from the Oxygen Minimum Zone which would have thrown off readings. 
• One scientist with whom we spoke said that the results from Sweetman’s work match almost exactly with data collected from landers which had accidentally trapped an air bubble. 
• Another scientist noted that if hydrolysis is occurring, we should see elevated levels of H from the reaction, and that would create higher levels of acidity in the solution, but that is not evident in any of the studies, including Sweetman’s.
• Finally, a number of people have noted the curious flattening of the oxygen production slope in Sweetman’s data (even declining in some cases).  Since this is an abiotic reaction, we would expect more of a continuous upward slope over the time period, even if the rate or slope varied over time.  The idea that there is a fixed quantity of oxygen within the lander, rather than a continuous reaction lends some support to the idea that there was an oxygen bubble present in the landers despite Sweetman’s assertion that this was not the case. 

What would Sweetman’s results mean in the context of nodule extraction?

While the scientific relevance of his results, vis a vis the origins of life, might be enormous, the quantities of oxygen being produced per Sweetman’s calculations are quite small.  In fact, they are orders of magnitude smaller than the quantity of oxygen produced, and CO2 captured, by trees in rainforests where we currently mine many of the same energy minerals today. 

Sweetman calculates in the paper the O2 production rate at b/w 1.7-18 mmol O2 m^-2 d^-1.  If we assume a relatively high rate of 12 mmol, that translates to 0.012 mole/day/m²  of O2 production.  This assumes a constant rate of production, even though Sweetman’s data shows that production actually stops and doesn’t appear to be a continuous process (and Sweetman concurs that continuous production is unlikely).  Regardless, at 32 grams/mole for O2, this means 0.384 grams/day/m².  That translates to 140 grams per year per sq meter or 0.31 lbs/year/m². 

Assuming that the nodules do, in fact, generate oxygen, and that they do so at a constant rate rather than the sporadic manner indicated in the paper, a square meter might produce a third of a pound of oxygen per year.  By contrast, a tree on average occupies around half of one sq meter at its base and produces approximately 260 lbs of oxygen per year.  It also absorbs around 50 lbs of co2.  https://www.thoughtco.com/how-much-oxygen-does-one-tree-produce-606785

We used the base of the tree because when we strip mine a m² of forest it takes down everything above.  Taking a sq meter of abyssal plain just doesn’t involve much damage on the vertical axis because there’s little there connected with the sediment and/or nodules.  If we take the 260lbs/0.5sqm (or 520lbs/sqm), and then we allow for spacing between trees such that they cover 19% of the area, we get a figure of 100 lbs/m² which would mean that we are destroying 100x the oxygen creating capacity when we strip mine in a rainforest vs. when we harvest nodules. Obviously, there are a number of assumptions in this exercise and none of them will be perfectly accurate, but it is quite clear from the analysis that oxygen generation in tropical rainforests is many times greater than on the abyssal plains (if it exists at all). 

Some will argue that even if the amount of oxygen created is small, this oxygen is still vital to the ecosystem functioning of the deep sea.  This could be so, however, it’s difficult to argue that it would be more consequential than the ecosystem functioning of tropical rainforests, and other forests destroyed by strip mining, given the much more consequential oxygen production in these environments.  In addition, we would point out that even if we were to harvest sufficient nodules to supply all of the cathode minerals required under the IEA’s Stated Policies Scenario over the next 20 years, we would only touch 0.06% of the abyssal plains (including plume impacts which are assumed to be limited per the latest empirical data).  So, the scale of impact is very small.   

Finally, if this reaction is happening, it’s happening in many places all over the world.  It’s not unique to nodules and its not unique to the abyssal plains.  It could even be occurring in terrestrial locations at the water table, or in lakes and rivers.  We should be rushing to gather these nodules (energy factories) to power our society if the hypothesis is correct, and we can rest assured that oxygen production from other rocks will remain unaffected. 

We remain skeptical of Sweetman’s hypothesis, but are open to reconsidering if the data is confirmed independently under proper methods.  In any event, we do not believe that this work represents a valid reason to accelerate the development of the nodule extraction industry.  There are many other reasons to pursue nodules, and this statement is backed by a substantial library of compelling science as we demonstrate at https://comrc.org/ .