- d + d → 4He + ɣ (23.8 MeV)
- p + d → 3He + ɣ (5.49 MeV)
The main reason these fusion reaction branches pose a difficulty is that branch (1) is believed to be involved in LENR in the context of PdD, somehow, through some process, whether directly or indirectly, and yet no gammas are seen. And (2) has recently come up as a possibility for LENR in NiH. People involved in trying to explain LENR have taken various roundabout ways to get around the problem of missing gammas, including Bose-Einstein condensates in which four or more deuterons fuse simultaneously.
The message to Vortex summarized points from an ongoing discussion I've had with Robin van Spaandonk concerning a proposal set out by Ron Maimon. Ron suggested in a post to the physics site physics.stackexchange.com in 2011 that what is going on in LENR is that the energy that would normally be emitted as a gamma in fusion reaction (1), above, is instead transformed into kinetic energy upon the decay of a short-lived [dd]* resonance that is created from the fusion of two deuterons in close proximity to a palladium lattice site, brought about in a very specific way. In broad terms, Ron proposes that the behavior of reaction (1) is different in close proximity to a palladium spectator nucleus than when it occurs in a plasma or in a vacuum. Because the electromagnetic dumping of the mass energy of the [dd]* intermediate state into a nearby source of electrostatic charge such as a palladium lattice site would be much faster than the emission of a gamma photon, it would be competitively favored. Since the process is also faster than the two other known dd fusion branches, it would compete favorably against them as well. The result would be a 4He daughter that pushes off of the palladium lattice site with 22.9 MeV of energy, instead of a prompt gamma, a triton and a proton, or a 3He and a neutron, products that are known from existing dd fusion experiments.
In the Vortex thread, Ron's idea has been taken and modified a little. I suspect that the transfer of energy via the electromagnetic interaction can also happen with the electronic structure of the palladium metal, far away from the lattice sites. If this were to happen, and there is reason to think it would be the dominant outcome, the momentum of the reaction would be absorbed either by an ensemble of electrons in the local system or by a single electron, and the daughter 4He would be almost motionless.
Somehow this disruption of the local system would then feed back into the reaction. I suspect that this occurs by a modification of the electronic structure of the palladium lattice; in fact, of that of any metal and not just palladium. The form of the disruption is perhaps a change in the charge density in the lattice. Normally most of the electrons in a metal are tightly bound around the lattice sites. The disruption would be to somehow modify the charge density so that some of the electron orbitals extended out into the interstitial areas, where the deuterium and hydrogen nuclei are. The electrostatic dumping of energy upon the decay of the [dd]* intermediate state would somehow propel this process forward. An early source of the idea that charge density might play a role in cold fusion was a paper written by a group at UC Berkeley in 1989, when a large number of people were trying to understand the results of Pons and Fleischmann. The Berkeley group do not appear to have taken into account the possibility that the charge density could be modified under certain conditions, however.
Getting the cold fusion process started would presumably require something other than fusion. I suspect in the case of Andrea Rossi's E-Cat and possibly Defkalion's Hyperion that this is through the agency of a thermionic emitter, which emits beta particles when heated up sufficiently. These devices could be using a compound such as lanthanum hexaboride, for example, whose properties are similar to those of modern lighter flint. As the heat of the substrate is increased by resistance heaters, or the thermionic emitter is stimulated by spark plugs, the compound would start to give off beta particles. Once that happened perhaps the charge density of the electronic structure of the metal lattice would be modified in some as-yet-unknown way, and dd and pd fusion would become many orders of magnitude more likely.
Ron Maimon does not like the idea of electron screening playing a role, for he believes it does not work at the scales involved in fusion (his explanation can be found in the comments here). But I suspect there is something like this going on, and that it can be effective, for the mechanism of the Polywell reactor works on a similar principle, and the explanation as a whole provides a very good phenomenological fit with the experimental data along a number of lines, including the observation that 4He appears to be born with almost no energy, as seen in a lack of Bremsstrahlung and prompt radiation.
So if modification of electron charge density or something like the mechanism behind the Polywell reactor is occurring, it might be possible to catalyze it using a thermionic emitter. What is interesting in this connection is that researchers have noted a possible connection (p. 219 ff) during electrolysis between oxide in the substrate and excess heat. Elsewhere I have learned that oxidation can sometimes lead to a lower work function, which is what is behind thermionic emission. This is, then, another detail that points in the direction of this general line of investigation.