Solid State Fusion & Radiochemistry

Introduction

The most discussed result in solid-state fusion is easy to state. In a set of experiments, palladium electrodes sitting in heavy water (water in which ordinary hydrogen is replaced by deuterium, a heavier form of hydrogen) gave off more heat than the electricity driving them could account for. The electrodes that ran hotter also ended up holding more helium-4. Roughly, the energy released per helium atom landed near the value you would expect if deuterium were fusing into helium.

Notice what kind of claim that is. It does not rest on a new theory of the nucleus. It rests on a measurement: can you show that a few helium atoms in every billion, made inside a sealed cell, are different from the helium already in the lab air? Air is about 5 parts per million helium, and helium slips through almost anything, including a rubber seal or a weld. So the real question is whether the helium you found came from the experiment or from the room. That is a measurement problem before it is anything else, and radiochemistry is the field built for exactly this kind of problem.

Why is this a measurement problem, not a physics problem?

Radiochemistry exists to pull a small, specific signal out of a noisy background and then defend the number.

Radiochemistry, stripped down, is the job of figuring out which specific kinds of atoms are in a sample and in what amount, usually when the amount is tiny and the background dwarfs it. Its standard equipment (blank samples that tell you what your instrument reads when nothing is there, and known tracers added on purpose) exists to separate a faint, specific signal from noise and contamination. A classic 1968 paper made the logic precise, drawing a line between the level at which you can fairly call something “detected” and the higher level at which you can put a trustworthy number on it.¹

Now line up the field’s main claims. Helium-4 building up in step with heat. Tritium above background. Occasional reports of elements that were not in the starting material. Each says a particular kind of atom is present at a particular level, and each is doubted on the same grounds: is the signal real, or is it the background wearing a disguise? The link between radiochemistry and solid-state fusion is not a loose analogy. The field’s hardest questions already are radiochemistry questions.

What can radiochemistry measure here?

The helium claim

Contamination weighed on the 2004 Department of Energy review, though it was not the only concern; reviewers also faulted experimental design and the missing neutron and gamma signals that a fusion heat source should carry.² On helium, the panel split, with several noting that some reported levels ran close enough to air that a slow leak could explain them. Proponents counter that some cells ran well above background. The radiochemist’s response is to measure it out of existence rather than win the argument by assertion. A noble-gas mass spectrometer can read helium far below the level in air, provided the sample is collected airtight and the whole line is checked with blanks. Better still, the ratio of helium-3 to helium-4 works as a fingerprint. Air carries a known ratio. Ordinary deuterium fusion would run rich in helium-3 and produce neutrons. The claimed reaction would produce almost no helium-3 at all. A single number for total helium can be mimicked by a leak. A ratio measured against a clean blank cannot, at least not easily.

Think of total helium as hearing a shout in a crowd and asking whether it was your friend. The ratio is like recognizing the specific voice. The comparison is loose: a voice is unique to one person, whereas an isotope ratio only points strongly, not with certainty.

The tritium claim

Tritium is a radioactive form of hydrogen, and detecting it is routine; you can count its faint radiation down to very low levels. The catch is subtler, and radiochemists know it well. An electrolysis cell concentrates tritium on its own, because splitting water preferentially removes the lighter hydrogen and leaves the heavier tritium behind. As the liquid volume shrinks, tritium climbs, with no fusion required. The correction is standard: calibrate how strongly the cell concentrates hydrogen isotopes, then ask whether there is more tritium than concentration alone can explain. The reported signals here are preliminary and unevenly reproduced. Whether they survive the correction is a question the field can settle cleanly.

The new-elements claim

This is the weakest of the three and the hardest to test. Elements sit at trace levels in the reagents and glassware, and in fingerprints, so a claim that a new element appeared is, by default, a claim about contamination. The check is the isotope ratio again: a genuine transmutation product should carry a ratio shifted away from the natural one, while ordinary contamination carries the natural ratio. That narrows the field without closing it, since processed material can arrive already shifted. Nobody has yet done the careful, independently repeated version with full accounting for blanks.

Underneath all three sits one accounting problem the field is unusually well equipped to close. If deuterium really is fusing fast enough to warm the cell, ordinary physics demands a specific set of products: two channels in roughly equal measure, one making tritium, the other making helium-3 plus a neutron, with the helium-4 channel suppressed by a factor of a million or more. The claimed pattern, heat and helium-4 with almost no neutrons or tritium, breaks that expected balance. That mismatch is the central reason physicists doubt the field. A complete inventory of what the cell produced (helium-4, helium-3, tritium, and any new elements, all measured against one shared set of blanks) would either confirm a real imbalance that demands new physics or dissolve it into ordinary backgrounds. No single instrument decides this. A coordinated count does.

What does the problem give back to radiochemistry?

The traffic runs both ways. Pulling helium and tritium out of a metal packed with hydrogen, close to one hydrogen atom for every metal atom, is a hard extraction problem, because the atoms you want are trapped in a solid that resists releasing them. Solving it sharpens methods that matter elsewhere, from dating rocks by their helium content to monitoring tritium in the environment. The controversy also funds instruments and trains people. In 2023 the U.S. agency ARPA-E committed about $10 million across eight teams, including groups at MIT, Stanford, and Lawrence Berkeley, to settle the question with modern methods.³ Whatever those teams find, the mass spectrometers and the analysts trained to run them carry over to the rest of nuclear measurement.

There is a larger prize, stated as the conditional it is. If the heat-and-helium pattern did survive a full assay, it would mean the surrounding environment could change how a nucleus divides its products, which would be a first-rank result. The reason to doubt it is exactly that the claim is so large, and a large claim sets a high bar rather than a low one.

So where does the evidence stand?

None of this asks anyone to believe the field’s claims. The strongest positive signal comes mostly from one community and is contested. The heat-and-helium correlation was reported principally by the chemist Melvin Miles, with supporting observations from SRI and ENEA, but largely within the field’s own conference literature, and the specific energy-per-helium figure has not been independently reproduced at high resolution in a mainstream journal.⁴ When a team organized by Google revisited the field with modern instruments and published in Nature in 2019, it did not reproduce the excess heat, and said so plainly, while also pointing to the materials science of extreme, sustained hydrogen loading as a real, underexplored corner where the older work was not good enough to settle anything.⁵ That is the real state of play. Not a proven phenomenon, and not a closed case. A poorly mapped region with a few persistent, unexplained signals in it.

The value radiochemistry brings does not depend on which way those signals resolve. The task is to make the measurements good enough to be believed either way, held to the ordinary standard of the discipline: characterized blanks, tracer methods, a clear rule for what counts as detected, and replication across independent labs. Applied properly, that standard can close the file as readily as open it.

Why would a curious student take this on?

The claimed effects are, at bottom, statements about which atoms are present and how many. Statements of that kind are radiochemistry’s to settle, and a field that can measure a fraction of a becquerel of tritium in a liter of water, or a trace of helium against the helium in the room, is not out of its depth. Bring the noble-gas line, the counter, and the discipline of blanks to a well-characterized cell, and one of two things happens. A signal survives to that standard, and the field has a real anomaly that needs explaining. Or it resolves into leaks, enrichment, or contamination, and the file closes on terms everyone can accept.

The decisive experiment can be named. An independent, high-resolution helium measurement of a working cell, reporting helium-4 together with the helium-3 to helium-4 ratio and full accounting for blanks, run by a group other than the one that first reported the correlation, would settle the helium question in a single measurement. It has not been done to modern standard. The helium in the room is about 5 parts per million. The open question is simply whether you can measure past it, and whether, once someone does, the signal is still there.

Notes

1. Lloyd A. Currie, “Limits for Qualitative Detection and Quantitative Determination: Application to Radiochemistry,” Analytical Chemistry 40, no. 3 (1968): 586–593. ↩
2. U.S. Department of Energy, Report of the Review of Low Energy Nuclear Reactions (Washington, DC: U.S. Department of Energy, 2004). ↩
3. Advanced Research Projects Agency–Energy (ARPA-E), “U.S. Department of Energy Announces $10 Million in Funding to Projects Studying Low-Energy Nuclear Reactions,” press release, February 17, 2023. ↩
4. Melvin H. Miles et al., “Correlation of Excess Power and Helium Production during D₂O and H₂O Electrolysis Using Palladium Cathodes,” Journal of Electroanalytical Chemistry 346 (1993): 99–117. ↩
5. Curtis P. Berlinguette et al., “Revisiting the Cold Case of Cold Fusion,” Nature 570 (2019): 45–51. ↩

Introduction

Start with the result the field itself regards as its strongest, and notice what kind of result it is. In a series of palladium and heavy-water cells, cathodes that produced more excess power accumulated more helium-4, and the energy-to-helium ratio landed within about an order of magnitude of the roughly 24 MeV per atom expected if deuterium were fusing to ⁴He. Whatever one makes of it, that claim does not stand or fall on nuclear theory. It stands or falls on a measurement: whether a few parts per billion of ⁴He generated inside a cell can be told apart from the 5.2 parts per million of ⁴He sitting in the surrounding air, ready to leak through any fitting, weld, or elastomer seal. That question is radioanalytical before it is anything else, and radiochemistry is the discipline built to answer it.

The Measurement Is the Claim

Radiochemistry, stripped to its function, is the quantitative determination of specific nuclides, usually at trace levels, usually against a background that dwarfs the signal. Its working furniture (procedural blanks, isotopic tracers and carriers, radiochemical yield determination, hermetic sampling, and a statistical account of what “detected” even means) exists because the central task is to pull a small, specific nuclear signal out of noise and contamination and defend the number afterward. The foundational 1968 treatment of detection limits in radiochemistry did that defending formally, separating the critical level at which you may call something detected from the detection limit you can expect to reach and the higher level at which you can actually quantify. Those are precisely the distinctions an excess-heat cell forces on you.

Now line up the field’s signature empirical claims. Helium-4 accumulating in proportion to heat. Tritium above background. Occasional reports of elements absent from the starting material. Each is an assertion that a particular nuclide is present at a particular level, and each is contested on the same grounds every hard radiochemical measurement is contested: is the signal real, or is it the background wearing a disguise? The connection from radiochemistry to SSF is not an analogy. The field’s hardest questions are already radiochemistry questions, whether or not anyone has labeled them that way.

What Radiochemistry Can Actually Measure

Take the helium claim first, because it is both the strongest positive signal and the cleanest illustration. Contamination weighed heavily in the 2004 Department of Energy review, though it was far from the only concern: reviewers also faulted experimental design, documentation, and background control, and pointed to the missing neutron and gamma signature that any deuteron-fusion heat source should carry. On helium specifically the panel divided, with several noting that some reported ⁴He levels ran close enough to atmospheric that a slow leak could account for them. That premise is itself disputed, since proponents report cells in which ⁴He ran well above background and occasionally above air, which skeptics in turn ascribe to leaks or handling. The radiochemical response is not to adjudicate that dispute by assertion but to instrument it out of existence. Noble-gas mass spectrometry measures ⁴He and ³He far below atmospheric levels, provided the sample is collected hermetically and the whole line is characterized with procedural blanks and isotope-dilution standards. The ³He/⁴He ratio is itself a fingerprint. Air carries a known, well-measured ratio. Ordinary D+D fusion, whose ³He+n branch accounts for about half of all events, would run ³He-rich and sit well above the atmospheric ratio; the claimed neutron-free ⁴He channel would do the reverse, producing almost no ³He and a ratio far below air. Either way the signature departs from atmospheric, and the near-absence of ³He is the specific mark of the anomalous channel. A number for total helium is ambiguous. An isotope ratio measured against a characterized blank is not. These tools are ordinary in noble-gas geochemistry labs; they have simply not always been brought to SSF cells with the discipline the problem demands.

Tritium sharpens the point, because here radiochemistry knows the trap better than anyone. Tritium is a pure low-energy beta emitter, half-life 12.32 years, endpoint near 18.6 keV, assayed for decades by liquid-scintillation counting down to fractions of a becquerel per liter. So the measurement itself is routine. The catch is that an electrolytic cell is, by construction, a tritium-enrichment device. Water electrolysis preferentially evolves the lighter hydrogen isotopes, so tritium concentrates in the residual electrolyte; dedicated rigs exploit exactly this to raise tritium levels by roughly one to two orders of magnitude before counting. A rising tritium concentration in a shrinking electrolyte volume is therefore the expected behavior of an ordinary cell, not a signature of production. Radiochemistry already corrects for this, calibrating the enrichment with the co-measured deuterium separation factor and asking whether the tritium inventory exceeds what enrichment alone can explain. The reported SSF tritium signals are preliminary and unevenly replicated, and should be read as reported rather than established. Whether any of them survive the enrichment correction is a question the discipline can settle cleanly.

The transmutation reports (new elements appearing where they were absent) are the weakest of the three claims and the most demanding to test, and they are where radiochemical contamination discipline stops being a nicety. Elements sit at parts per billion in reagents, electrodes, glassware, and fingerprints, so a claim that a new element appeared is, by default, a claim about a blank. Neutron activation analysis and isotope-ratio and inductively-coupled-plasma mass spectrometry can quantify trace elements and, more tellingly, their isotopic signatures: a genuine transmutation product should carry a ratio shifted from the natural one, whereas ordinary contamination carries the natural ratio. That test is necessary but not sufficient, since processed or prior-lab material can arrive already isotopically shifted; a shifted ratio narrows the field rather than closing it. These claims remain preliminary and poorly replicated, and belong among hypotheses still to be tested to radiochemical standard. Independent activation-analysis replication with full blank accounting is the price of admission, and it has not been paid.

Underneath all three is a single accounting problem the discipline is unusually equipped to close. If deuterons are fusing fast enough to warm a calorimeter, the ordinary D+D channels demand a specific ledger: roughly equal branches to tritium-plus-proton and helium-3-plus-neutron, with the ⁴He-plus-gamma channel suppressed by six or seven orders of magnitude. The claimed pattern, heat and ⁴He with almost no neutrons or tritium, violates that ledger, and that imbalance is the central nuclear objection to the field. A complete radiochemical assay of the ash, quantifying ⁴He, ³He, tritium, and any activation products together against a common set of blanks, is the measurement that would either confirm the anomaly as a real imbalance demanding new nuclear physics or dissolve it into ordinary backgrounds. No single instrument decides this. A coordinated nuclide inventory does.

What SSF Returns to Radiochemistry

The traffic runs both ways, and the return cargo is more concrete than a courtesy. Extracting and quantifying helium and tritium from a metal lattice loaded with hydrogen to near one-to-one is a genuinely hard matrix problem, because the analyte is trapped in a solid that resists degassing and whose composition fights the separation. Solving it pushes noble-gas extraction and low-level tritium assay into territory that spills directly into helium thermochronology, environmental tritium monitoring, and fusion fuel-cycle accounting, each of which needs better ultra-trace helium and tritium metrology than it has. A loaded cathode is, for the analyst, a demanding unknown, and demanding unknowns are how methods improve.

There is a larger prize, stated as the conditional it is. If the heat-helium pattern survived a full radiochemical assay (a real if), then an environmental influence on nuclear branching ratios would be a result of the first rank, and squarely a nuclear-chemistry result, since branching is exactly what the discipline measures. The prior against it is low precisely because the claim is large, and that sets the evidentiary bar high rather than lowering it. The more modest returns are more certain. The controversy already funds instruments and trains people: in 2023 ARPA-E committed roughly $10 million across eight teams, including groups at MIT, Stanford, and Lawrence Berkeley, to settle the low-energy-nuclear-reaction question with modern methods. Whatever those programs find, the mass spectrometers, the calorimetry standards, and the trained analysts who run them transfer to the rest of nuclear measurement science.

The Honest State of the Evidence

None of this asks radiochemistry to believe the field’s claims. The strongest positive signals are single-community and contested: the heat-helium correlation has been reported principally by Miles, with supporting observations from SRI and ENEA in helium-leak-tight calorimeters, but largely within the condensed-matter-nuclear-science conference literature, and the quantitative ~24 MeV ratio has not been independently replicated at high resolution in a mainstream venue. It has also been criticized on calorimetry and contamination grounds. The tritium and neutron reports are preliminary and unevenly replicated. When a Google-convened team revisited the field with modern instrumentation in the late 2010s, it did not reproduce excess heat and said so in Nature, while also identifying the materials science of extreme, sustained hydrogen loading as a genuinely underexplored corner where the older work was not good enough to conclude anything. That is the actual state of play: not a proven phenomenon, not a closed case, a poorly mapped region with a few persistent, unexplained signals in it.

The value radiochemistry brings does not depend on which way those signals resolve. The task here is to make the measurements good enough to be believed either way, and the standard for that is the ordinary standard of the discipline: characterized blanks, isotope-dilution and tracer methods, the Currie framework for what counts as detected, and replication across independent labs. Applied honestly, that standard is as capable of closing the file as of opening it, and both are worth doing.

Why Take This On

The claimed phenomena are, at bottom, statements about which nuclides are present and in what amounts. Statements of that kind are radiochemistry’s to adjudicate, and a discipline that can measure a fraction of a becquerel of tritium in a liter of water, or a trace of ⁴He against the helium in the room, is not out of its depth here. Bring the noble-gas line, the scintillation counter, the activation analysis, and the blank discipline to well-characterized cells, and one of two things happens. A signal survives to radiochemical standard, and the field has a real anomaly that demands explanation. Or it resolves into leaks, enrichment, and contamination, and the file closes on terms everyone can accept. The decisive experiment is nameable. An independent, high-resolution noble-gas assay of a working cell, reporting ⁴He together with the ³He/⁴He ratio and full procedural-blank accounting, run by a group other than the one that first reported the correlation, would settle the helium question in a single measurement. It has not been done to modern standard. That is the test the rest of this argument has been circling. The helium in the room is 5.2 parts per million. The whole question is whether you can measure past it, and radiochemists measure past worse every day.

Editorial note: This article presents a scholarly synthesis of SSF’s relationship to radiochemistry. The underlying nuclear claims of SSF/LENR remain scientifically contested. Evidence claims are tiered as established, contested, or reported-but-unconfirmed as noted inline. Readers are directed to primary experimental literature for empirical evaluation.