The published results from experiments similar to the ones described here show a zero-signal cutoff in Kurie plots at 510–520 keV close to the electron mass of 511 keV [3–5, 26], indicating electrons and/or positrons as the particles formed in large numbers by the detection process. The initial suggestion [3, 4] to explain this apparent beta decay process was that the muon decay gave rise to this signal. When other experimental features were observed, like the strong increase in signal by using glass or metal converters [5], a simple muon decay scheme was clearly no longer a possible explanation. A pair production of the observed type requires apparently bosons with high energy.
A mechanism which agrees with experimental findings indeed exists.The muons have high energy of the order of 100 MeV and give gammas and x-ray photons in interaction with matter for example in the converter. These photons then create numerous e+e− pairs in materials around the detector and these pairs are detected in the PMT.
4.1. Constant zero-signal cut-off
The shape of the MCA energy spectrum and its high-energy cutoff, which indicates pair production, is unchanged during changes in source intensity, as shown in Fig. 3. Numerous spectra have also been published which show that the cut-off is constant close to 511 keV [3–5]. Such experiments show the intensity variation of the source with several source parameters. Another straight-forward experiment is to change the distance between the converter where the e± pairs are formed and the PMT detector. Such an experiment is shown in Fig. 4. The conditions for both panels in Fig. 4 are the same, with the bottom panel measured after a 10 min rebuilding time, moving the PMT 10 cm away from the steel plate giving 100 times lower signal but still the same signal cut-off.
With the PMT close to the stainless steel plate in the vacuum flange in front of the detector part, the cutoff is observed in the Kurie plot to be at around 180 channels or 540 keV. This slightly higher value than in most experiments is apparently, as seen in the upper panel of Fig. 4, due to a higher background intensity from another process which has its cutoff at around 1 MeV. This could be a slightly different process where complete lepton pairs i.e positronium at 1.02 MeV total energy are detected. This cutoff is then due to the maximum positronium kinetic energy of 1.02 MeV in the energy distribution. At a large converter distance as in the bottom panel in Fig. 4, these positroniums have annihilated giving only gamma radiation, with the known lifetime for the singlet state of 0.12 ns and of the triplet state of 142 ns [28, 36].
The published spectra contain several examples of the constant cut-off for different converter materials [3–5]. That the cut-off is independent of the converter material supports a muon-gamma-e± pair model. The nuclei in the converters do not influence or take part in this process.
4.2. Muons are not ambient but from the generator
The signal in the experiments varies in size with many different parameters. One example is shown in Fig. 5 where the signal is shown to depend strongly on the amount of D(0), being deposited in the source by gas admission prior to the experiment and then removed by heating in vacuum. Renewed gas admission increases the signal again. This process is reproducible and is used routinely. This proves that the muons originate in the source and are not ambient, for example from nuclear processes in the upper atmosphere as most mouns on Earth are.
The PMT-MCA signal in the region above 1 MeV is intense and complex. An example of the signal with the detector mounted on the chamber and also separate from the chamber is shown in Fig. 6. The low-energy muon part below 1 MeV is almost unchanged when the detector is moved away from the chamber. The high-energy part is however strongly reduced at long distance, suggested to be so since it is caused by the more shortlived kaons and pions from the annihilation reactions [10].
4.3. Function of the converter
The importance of the converter is demonstrated by an experiment in Fig. 7. At 1 MeV energy, the signal is increased a factor of 1000 by the converter plus NaI snintillator in front of the PMT, and at 500 keV about a factor of 20. The Al converter is placed between the scintillator and the PMT, thus no visible photons from the scintillator can reach the PMT.The intensity at 2.5 MeV in Fig. 7 is not changed by the converters. This behaviour agrees with that expected for muons at low energy. Kaons and pions, which are not influenced strongly by the converters, give the signal at high energy as mentioned above.
The best Al converter is a compressed “pillow” of a few layers of 20 µm thick Al foil. In the results shown in Fig. 8, it should be noted that also the process of transport of the final e± leptons to the PMT is of importance for the size of the observed signal. A dense converter material like Ta will give less penetration of the e± particles relative to glass and Al. Another drastic example of this transport effect is shown in Fig. 9. The use of two plastic scintillators of length 5 cm in front of the converter gives a signal which is 10–100 times lower than without the scintillators. The non-conducting organic scintillator material prevents the transport of the electrons and positrons to the PMT.
4.4. Lifetime of H(0)
The spontaneous nuclear process in H(0) implies a lifetime of H(0). So we need to estimate this lifetime from the results found here. In Fig. 4 upper panel the total intensity of muons is approximately 105 particles per 200 channels, thus 2×107 particles during 500 s or 4×104 s− 1. The total volume of H(0) on the generator is estimated to be a 4 mm diameter layer with a thickness of 1 µm or a volume of 6×10− 12 m3. With a bond distance of 2 pm in spin state s = 2, each atom in H(0) occupies a volume of 8×10− 36 m3. This means 8×1023 atoms in the layer on the muon generator which produces 4×104 reactions s− 1. So the reaction rate per atom becomes 5×10− 20 s− 1 or a life time of 1011 y thus longer than the assumed lifetime of the universe but much shorter than the lifetime of a proton outside H(0) which is thought to be > at least 1032 y [37]. The spontaneous nuclear processes giving the mesons and the muons are probably the same as for the laser-induced nuclear orocesses, thus baryon annihilation [6, 7]. This process is initiated by a transfer in H(0) from spin state s = 2 to s = 1 creating mesons, muons and finally only gamma photons, electrons (positrons) and neutrinos. Thus a mechanism exists for proton ”decay” to mesons and muons.
4.5. Statistical energy distributions
A statistical theory was described above which gives energy distributions of a few kinetic energy terms out of larger number of terms, assuming a fixed total energy to be distributed among this larger number of energy terms. This is an energy distribution for a so called microcanonical ensemble of systems. An energy distribution which is peaked at an intermediate energy is found for two terms out of for example six terms (N = 6). An even better agreement is here found with N = 5.5. Such a distribution is shown in Fig. 10, with the total energy limit put equal to 175 channels. In this way, a simple comparison is possible with the normal MCA plots. This type of distribution is similar to a beta-decay distribution [35], which also has a fixed total energy to distribute between several particles and thus between several energy terms.
For an accurate comparison with the Kurie plots shown in Figs. 3 and 4 and in previous publications [3–5], the square root of the statistical energy distribution is also calculated. The best result giving linear plots for this (P(ε))½ function is shown in Fig. ,10 for this case N = 5.5.
One might expect the best agreement for N = 6, thus six quadratic energy terms in total. Two velocity vectors are involved in each pair production step (one electron and one positron). Each such vector is associated with three quadratic kinetic energy terms for vx, vy and vz respectively. However, they are not all independent since linear momentum needs to be conserved for each electron and positron pair. Such restrictions often give lower values for the apparent number of terms in this type of statistical energy distribution [33]. The energy distributions observed may be thought to correspond to three terms. However, the third term will tend to deflect the particle away from the detector, so two terms for the observed energy is more correct. Thus, the observed energy distributions should correspond to two terms out of 5 or 6. Here, the plot in Fig. 10 shows a linear Kurie plot to the theoretical cut-off introduced in the statistical formulas, thus giving agreement with experimental results for two terms out of 5.5.
An experimental energy distribution is included in Fig. 11, using a theoretical distribution with N = 5. At high energy, the tail of the experimental distribution does not go to zero since another type of signal is found there as seen also in Figs. 3 and 4 which was described above.. In the range of 40–150 channels, the two curves agree quite well.
The observed particle energy is thus distributed to two terms out of five - six for an electron or a positron with a maximum energy of 511 keV. This indicates that the pair-production process is sequential for each pair produced. How energy is conserved in an annihilation process which generates many pairs simultaneously does not seem to be well understood, probably since such processes have not been investigated. Another annihilation process with several different particle pairs formed was recently studied, showing excellent energy conservation [6, 7].
The linear Kurie plots shown here and previously [3–5] are thus interpreted as due to e± pair-production.