Posted on by maxfomitchev

I decided to replicate an experiment described by Mark Prelas in his Neutron Emission from Cryogenically Cooled Metals Under Thermal Shock paper. In this experiment titanium shavings were cryogenically loaded with deuterium by cooling in liquid nitrogen. Then the cold chips were thermally shocked by injecting hot water. This thermal shocking / rapid heating reportedly produced a neutron flux in the millions of neutrons. Prelas of course was drawing on similar experiments performed by Menlove, where the above background neutron flux was reportedly observed when cryogenically loaded titanium sponge chips were allowed to warm up.

This experiment was easy enough for me to try. The experimental setup is shown on Fig. 1.

Fig. 1. The experimental setup for cryogenic loading of titanium shavings with deuterium.

For cooling I used a LN2 Dewar from my HPGe detector. I prepared titanium shavings by dry-turning a titanium rod. The outer layer shavings were discarded, only clean non-oxidized inner shavings were used. I deposited the titanium shavings into an aluminum cylinder (lecture bottle) immediately after the shavings were made, connected the cylinder to a small vacuum manifold with a port for a D2 regulator / cylinder, a Pfeiffer AR pressure gauge capable of reading up to 5000 Torr and a valve for evacuation. I evacuated this system to better than 1E-4 Torr prior to introducing deuterium at 3000 Torr. Then I dipped the cylinder into the LN2 Dewar. The pressure in the cylinder was dropping as the cylinder was cooling. An exothermic deuterium loading event was clearly observed by a spike in LN2 boiling during the cooling process (e.g. midway through the cooling process the cylinder was producing a burst of heat, which in turn created a vigorous LN2 boiling). Generally I waited for 30-40 minutes for the cylinder to cool and titanium shavings to load with D2. The final pressure in the cylinder was 760-801 Torr.

The neutron counting setup is shown on Fig. 2.

Fig. 2. Neutron counting setup.

I used an ANL-8 to capture counts from 8 helium-3 filled SI-19N proportional counters. The counters were arranged in a semi-circle and embedded in HDPE moderator. The aluminum cylinder was placed in a plastic jar, which I filled with boiling water to thermally shock the chips. Adding hot water resulted in pressure increasing in the cylinder to 2000 Torr (some deuterium must have exited the chips in the process). The neutron counts are shown on Fig. 3.

Fig. 3. Background neutron counts (red) vs. counts from the loaded titanium shavings (blue).

The neutron counts from the thermally shocked deuterium-loaded titanium shavings appear to be 13% higher than the background at an enormously statistically significant level (P = 0.001). But does this really mean that the observed 13% excess is real? On one hand yes – the observed increase in neutron counts is real. Those are really neutrons (judging by the shape of the thermal neutron spectrum), and there are more of them now. But the real question – are these neutrons really originating from the titanium shavings or is the increase is caused by a systematic error? To rule out the latter I have obtained a lot more neutron counts from the shavings (there are the ‘experiment’ counts), then removed the cylinder from the vicinity of the detector bank and obtained a few more ‘background’ counts. The results are shown on Fig. 4.

Fig. 4. Counts from the shavings (blue) vs. background neutron counts (red).

As one can see from Fig. 4. there is no difference in counts when the cylinder with titanium shavings is removed from the detector bank. This can mean only one thing: the counts on Fig. 3 increased for a reason not related to the presence of the deuterium-loaded titanium shavings, as the counts on Fig. 4. clearly indicate.

Conclusion

The thermally shocked titanium shavings in this experiment did not produce neutron counts in excess of the background.