Project Callisto

Forty-plus attempts did not reproduce the 1989 cold-fusion result

Callisto performed more than 40 formal and modified replications of the Fleischmann–Pons experiments, including prolonged electrolysis, cutting-fluid additives, oxygen bubbling, and high-input-power operation. Electrolyte levels and calorimeter thermal stability were more tightly controlled, although missing details in the original publications prevented an exact reconstruction of the 1989 experiments.

No excess heat was observed.

Full deuterium loading was achieved and measured in macroscopic pressure systems

High fractional loadings (above 90%) are theorized to be necessary for AHE. To fully exhaust the loading hypothesis, we needed to confirm stoichiometric, or 100%, loading. Reaching high loadings with gas requires extremely large pressures: a deuterium pressure of 70 ksi achieves approximately 90% loading, while increasing this to 95% requires almost twice as high a pressure, 135 ksi. At room temperature, approximately 400 ksi was not enough for full PdD loading; pressures above 450 ksi were needed. We built pressure systems capable of operating above this threshold so we could study fully loaded palladium.

Callisto made 150,000 psi a routine experimental environment and developed advanced cells reaching 600,000 psi. These were macroscopic pressure systems rather than diamond-anvil cells: they provided working volumes for samples, stimulation equipment, and diagnostics.

Carefully characterizing a sample’s loading level is essential to investigate AHE. We developed two direct loading-measurement techniques, UltraSorb and MicroSorb, as well as indirect calibrated measurements based on strain, resistance, and reflectance.

Full loading was reached through cryogenic experiments at 150 ksi and −100 °C. Room-temperature gas loading required pressures above approximately 450 ksi, which were maintained for extended durations.

No excess heat was observed.

The leading post-1989 LENR claims did not reproduce either

Callisto tested later claims by McKubre, Staker, Hubler, Storms, Miles, Boss, Barham, Energetics, and Letts. Some experiments were replicated exactly, down to sourcing vintage components based on correspondence with the original scientists. In other cases, critical elements were replicated, but existing infrastructure, such as sample holders and calorimeters, was used.

In a 60-day Staker-like experiment, palladium loading remained above 80% for nearly all cells throughout the experiment. The steady-state ratio of measured heat output to electrical input remained near one. Apparent excess-heat excursions were explained by energy accounting and chemical recombination, not nuclear reactions.

No excess heat was observed.

Precision calorimetry found no excess heat at ultrapressure, from −150 °C to 200 °C

Callisto measured heat released or absorbed by gas-loaded palladium using differential calorimeters, comparing deuterium-loaded samples with hydrogen-loaded controls. Samples ranged from nanoparticles to bulk foils. Across the gas-calorimetry campaign, deuterium pressures reached 150,000 psi and temperatures ranged from −150 °C to 200 °C. Loadings up to unity were achieved, and the calorimeter system was monitored at equilibrium for 3–14 days in each experiment.

In a separate cryogenic experiment, gas-loaded palladium was held at 130,000 psi and −100 °C while differential calorimetry monitored its heat flow.

No excess heat was observed.

More than 200,000 sample-hours of calorimetry found no excess heat

One possible explanation for unsuccessful experiments was that anomalous heat occurred only rarely. To explore this possibility, Callisto developed a system of 100 electrochemical cells with in-situ calorimetry running in parallel. Its first long-duration campaign ran 80 cells for 50 days, accumulating approximately 11 cell-years of calorimetry in a single experiment.

Across the electrochemistry program, we investigated more than 500 samples and performed more than 200,000 sample-hours of calorimetry. Alongside replications, we designed new experiments to test proposed requirements for AHE: higher loading, different temperatures, surface coatings, additives, electrical and mechanical stimulation, single crystals, and palladium alloys.

No excess heat was observed.

Thermal, laser, terahertz, and electric-field stimulation produced no evidence of AHE

Callisto tested whether added energy could initiate LENR in highly loaded palladium. We heated samples to melting, applied laser pulses on the fS timescale—faster than atomic motion—and tested intense electric fields. Microfabricated devices combined stimulation and temperature sensing on the same chip; the NanoTI campaign collected data from more than 448 devices over more than 1.5 million pulses, including tests above 450,000 psi.

In the laser experiment shown here, one pulse stimulated palladium while a second pulse measured the thermal response of an adjacent gold layer. At 145,000 psi, this measurement showed that an apparent difference between hydrogen- and deuterium-loaded palladium came from their different optical responses, not anomalous heat.

We also developed intense terahertz pulses and microscopic resonators to try to drive vibrations of deuterium within the palladium lattice. Metallic screening prevented useful coupling to the intended vibrations, limiting what this approach could test.

No excess heat was observed.

Searches for even a single reaction event found no evidence of AHE

Callisto developed detectors intended to record even a single LENR event. Palladium films were made thin enough that the energy from a hypothesized reaction would leave a permanent hole, visible in electron microscopy. Comparing the same areas before and after an experiment allowed us to distinguish new damage from pre-existing defects and contamination.

We used these detectors across experiments involving laser stimulation, intense electric fields, high temperatures, cryogenic conditions, and pressure-assisted electrochemical loading. Early nanopillar arrays were vulnerable to debris and handling damage, so we developed more robust nanofilms that made new features easier to identify and investigate. Across these tests, no features attributable to LENR were found.

Other experiments searched for brief infrared hot spots, persistent changes in magnetic sensors, acoustic impulses, and nuclear radiation. None established evidence of LENR; some detection methods remained under development when the program ended.

No excess heat was observed.

None of the apparent anomalies established evidence of AHE

Callisto repeatedly observed signals that initially looked like excess heat. In the electrochemistry program, every apparent excess-heat event was traced to a conventional explanation. Causes included errors in electrical-power accounting, changing calorimeter response, the different reaction entropies of hydriding and hydrogen evolution, faulty sensors, and stored chemical energy released when hydrogen and oxygen recombined.

The pressure program traced apparent anomalies to differences in gas cooling and optical response, temperature-measurement errors, contamination, and damage from laser stimulation or sample handling. Hydrogen controls, independent measurements, and before-and-after microscopy helped distinguish these effects from evidence of LENR.

A small number of nonrepeating observations in the pressure program remained unresolved. They did not establish LENR. The reports distinguish these observations from effects for which a cause was demonstrated.

No excess heat was observed.