Sunday, 27 September 2009

Resonating dark matter

On this blog I regularly follow the progress in dark matter building. One reason is that next-to-nothing is happening on the collider front: Tevatron invariably confirms the standard model predictions up to a few pathetic 2 point null sigma bleeps now and then. In these grim times particle theorists sit entrenched inside their old models waiting for the imminent LHC assault. The dark matter industry, on the other hand, enjoys a flood of exciting experimental data, including a number of puzzling results that might be hints of new physics.

One of these puzzles - the anomalous modulation signal reported by DAMA - continues to inspire theorists. It is a challenge to reconcile DAMA with null results from other experiment, and any model attempting that has to go beyond the simple picture of elastic scattering of dark matter on nuclei. The most plausible proposal so far is the so-called inelastic dark matter. Last week a new idea entered the market under the name of resonant dark matter. Since this blog warmly embraces all sorts of resonances I couldn't miss the opportunity to share a few words about it.

In the resonant dark matter scenario the dark matter particle is a part of a larger multiplet that transforms under weak SU(2). This means that the dark matter particle (who as usual is electrically neutral) has partners of approximately the same mass that carry an electric charge. Quantum effects split the masses of charged and neutral particles making the charged guys a bit heavier (this is completely analogous to the $\pi_+ - \pi_0$ mass splitting in the standard model). Most naturally, that splitting would be of order 100 MeV; some theoretical hocus-pocus is needed to lower it down to 10 MeV (otherwise the splitting is to large compared to nuclear scales, and the idea cannot be implemented in practice), which is presumably the weakest point in this construction.

Now, when dark matter particles scatter on nuclei in a detector there is a possibility of forming a narrow bound state of the charged partner with an excited state of the nucleus, see the picture. That would imply that the scattering cross-section sharply peaks at a certain velocity corresponding to the resonance.The existence of the resonance is very sensitive to many nuclear parameters: mass, charge, atomic number and the energies of excitation levels. It is conceivable that the resonant enhancement occurs only for one target, say, iodine present in DAMA's sodium-iodine crystals, while it is absent for other targets like germanium, silicon, xenon etc. that are employed in other dark matter experiments. For example, the resonant velocity for these other elements might be outside the range of velocities of dark matter in our galaxy (the escape velocity is some 500 km/s so that there is an upper limit to scattering velocities).

So, that looks like a perfect hideaway for DAMA, as other experiment would have a hard time exclude the resonant dark matter hypothesis in a model independent way. Fortunately, there is another ongoing dark matter experiment involving iodine: the Korean KIMS based on cesium-iodine crystals. After one year of data taking the results from KIMS combined with those from DAMA put some constraints on the allowed values of the dark matter mass, the position of the resonance and its width, but they leave large chunks of allowed parameter space. More data from KIMS will surely shed more light on the resonant dark matter scenario.

3 comments:

  1. So what does this resonance do to their efficiency for detecting the 3.2 keV Auger photon from 40 K decay? Something has to suppress the 40 K decay in the region around 3.2 keV as well as the continuum background, since the majority of hypothetical DM events do not contribute to the oscillation, but instead are supposed to contribute to this peak at 3.2 keV or flat background there, which just magically happens to be in the same place as the 40 K transition. See arXiv:0804.2738 especially p.15 and Fig.9. Then take a look at Fig.26 and you will notice their efficiency for detecting the 3.2 keV photon is only about 55%.

    So DAMA is more likely explained by 40 K decay in which the primary 1.5 MeV decay is missed by their detector (for instance occurs near a surface in which the beta escapes, or occurs in the EC branch of the 40 K decay so that there is no beta in the final state). Oscillation must be explained by a ~2% time variation on their 55% efficiency. They do not evaluate the time variation of their efficiency.

    This hypothesis can be confirmed by looking in the unpublished low energy bins (< 2 keV) which should oscillate even more than the bins in 2-4 keV. (Assuming detector noise has the same oscillation properties as their efficiency)

    ReplyDelete
  2. Definitely, 40K is the main culprit here. But the matter is far from settled: it is not clear why their efficiency should peak in June, and there is no modulation observed in double hits. In any case, I've written so much about wild DAMA speculations that I probably owe a separate post about down-to-earth explanations.

    As for the rDM, they assume that the resonance velocity is at the Maxwell tail, which boosts modulated vs unmodulated. I dont think they subtract 40K when they derive their boudns from unmodulated DAMA rate; probably there is no way to estimate that.

    ReplyDelete
  3. Why does the 40 K peak oscillate? Who cares? It's 40 K, not some exotic particle that conspires to place a peak right on top of a known background.

    What really killed DAMA for me was arXiv:0808.3283, "Evidence for Correlations Between Nuclear Decay Rates and Earth-Sun Distance". This paper has 3 isotopes which oscillate. Some are alpha emitters, some beta (so a common physics cause is unlikely). I dug into the original source papers and found the causes of most of these oscillations, and they're all of experimental origin, and they say so in the original papers. (The originals do NOT claim to observe oscillation of half-lives as they know it is a systematic experimental effect) For instance one experiment has an air gap between the source and the detector, which introduces pressure, temperature, and humidity corrections that obviously have an annual variation. This paper itself is irresponsible, ignoring the explanations given in the source papers and proposing an exotic explanation for old measurements that were not performed by the authors.

    EVERYTHING has an annual modulation that peaks in June. Temperature, pressure, humidity, voltage, Berlusconi's philandering, etc. Because the efficiency for detecting this photon is only 55%, it is most likely something which affects the efficiency. For instance, variations in the ground voltage affecting the noise in photodetectors...

    For an experiment that claims to measure a time variation, DAMA does surprising little evaluation of the temporal stability of their equipment. (The temporal stability of their efficiency is not evaluated)

    Anyway, I shouldn't complain, all this attention on DAMA have left me alone with the really interesting problems...

    ReplyDelete