Wednesday, October 30, 2013

New Direct Detection Limits On Dark Matter From LUX

The vertical axis is cross-section of interaction, the horizontal axis is mass in GeV/c^2.
LUX pretty definitively rules out the possibility, hinted at by several dark matter experiments, of a dark matter particle in the 5 – 20 GeV/c² mass range. While XENON100 seemed to contradict this possibility already, it didn’t do so by a huge factor, so there were questions raised as to whether their result was convincing. But the sort of ~10 GeV/c² dark matter that people were talking about is ruled out by LUX by such a large factor that finding ways around their result seems nigh impossible. . . . by 2015 their results should improve by another factor of 5 or so.

From here (some parenthetical matter omitted). Lubos concurs with Matt Strassler's analysis above. He states:

The exclusion and the agreement with the background expectations [in the LUX experiment] is spectacular (no phantom events appear in their data at all) and I have no doubts that all the "Yes signals" [in other direct dark matter detection experiments] are due to some misunderstood background. This result hasn't excluded all meaningful models of dark matter yet but upgrades of this experiment are able to get to the region where the cross sections start to become unnaturally small.

Physics blogger Jester is also convinced.

Direct dark matter detection experiments already ruled out heavier dark matter particles up to about 1000 GeV/c^2.


This result confirms indirect astronomy inferences that rule out WIMP dark matter as heavy as 5 GeV or more, i.e. Cold Dark Matter, while not ruling out profoundly lighter "Warm Dark Matter" particles of ca. 2 keV (each of which would be about 2,500,000 times lighter).

The exclusion ranges rule out cross-sections of interactions comparable to those of neutrinos, (also here) which would be an a priori expectation for a weakly interacting dark matter particle.

Of course, the theoretical extreme of "collisionless" dark matter would be impossible to detect, by definition, in an experiment like LUX. Even an almost collisionless dark matter particle like a "sterile neutrino" which does not interact via anything other than gravity and Fermi contact interactions, would likewise have such a tiny cross section of interaction that this kind of experiment could probably not detect these particles.

This exclusion also highly constrains the mass of any lightest or next to lightest superpartner in a supersymmetry theory that could be a dark matter candidate, effectively closing a significant share of SUSY parameter space. The LHC rules out light superpartners up to the 100s or 1000s of GeVs. LUX rules out SUSY WIMPs of 5 GeV to about 1000 GeV. SUSY models with, e.g., a stable (or nearly stable) very light gravitino, in which all other superpartners are very heavy, are increasingly hard to construct, particularly if they are at all "natural."

We have long known that "hot dark matter" with particles of mass on the order of 1 eV or less, are excluded by astronomy data. Thus, dark matter has to be heavier than the known neutrinos. Taken together with this direct detection exclusion, the range of permissible masses for dark matter particles, if they exist, has to be of the same order of magnitude as already known fundamental fermions (other than the top quark) and the known hadrons - probably at the very low end of this range.


Maju said...
This comment has been removed by the author.
Mitchell said...

At the moment I feel that a right-handed neutrino would be the most conservative explanation for keV-mass warm dark matter. It would be one more feature of the neutrino sector (along with the mixings and the left-handed masses) that required a deeper explanation.

andrew said...


A particle that makes the astronomy data that motivates the search for dark matter has to be completely or almost "collisionless", i.e. capable of going through lots of seemingly solid matter without interacting with it. In contrast, atoms in interstellar gas constantly bump into each other and has a tendency to clump together.

Neutrinos are the only kind of matter known to behave like this that has been discovered, but that doesn't mean that there couldn't be something else out there that s similar but heavier. We know that neutrinos can't make up much of "dark matter" because they are extremely light and it is possible to accurately estimate how many of them there are in the universe and it isn't close. Also, if the universe's dark matter were made of neutrinos, there would be just a sea of stars without galaxies or galactic clusters in it (based on detailed computer simulations of that possibility).

In order to rule out background noise of particle that aren't collisionless the LUX experiment detectors are set up so that no ordinary atom or molecule could penetrate the detection chamber in this experiment deep in a South Dakota mine in a shielded tank of ultrapure water. Ordinary atoms and molecules are excluded by hypothesis and experimental design, except, for example, for a tiny number of atoms that are shed by the detector walls themselves (a number that they can calculate in advance that has certain characteristic levels of energy and directions which are removed from the raw detector data results as expected background noise).

The experiment was designed to be sensitive only to dark matter particles above a 5 GeV/c^2 low end cutoff, in part, because almost no hadron (i.e. composite particle made up of quarks akin to a proton or neutron) can be much heavier than that, so that anything that is detected must be "new physics", and so there isn't much of a background to screen out (and also because the smaller the mass of the dark matter particle, the harder it is to detect period, it requires a much bigger, more carefully calibrated detection setup and more accurate estimates of the background noise).

Ordinary matter that is not luminous, which I often call "dim matter" at this blog, is known to be out there and is surely undercounted somewhat.

But, if gravity behaves according to General Relativity and if one makes a handful of other very basic assumptions (e.g. that almost no particles with half-lives of a fraction of a second created only in superhigh energy collisions exist in nature, and if protons combine with each other in highly predictable proportions matched by experimental measurements of light from other suns), it is possible to calculate the total number of protons and neutrons in the universe, and hence the total amount of ordinary atoms and molecules out there. But, there aren't nearly enough atoms in the universe to provide all observed dark matter (total amounts of matter can be measured indirectly using techniques like observing how much light is bent by gravity in its vicinity without knowing what kind of matter it is).

If the dark matter paradigm is right (as opposed to a modification of gravity, or unknown, non-luminous structure of galaxies not in halo that arises dynamically in the same way in each galaxy for which we have almost no experimental hints), then dark matter has to be made out of something with properties very similar to neutrinos, but heavier that is not a proton or neutron or anything else made out of quarks or electrons or muons or tauons, or neutrinos.

The LUX experiment establishes that any particle that has those properties can't weigh much more than helium atom. At a level of classification, these particles would probably be akin to electrons or quarks, not to atoms or molecules.

andrew said...

@Maju (continued)

Other astronomy data strongly suggests that if there is dark matter at all, it should have a mass that is about 2000 times lighter than an electron and about 2000+ times heavier than neutrino, i.e. about 2000-3000 eV/c^2. So, the LUX data isn't really so much of a surprise, but puts another nail in the coffin of dark matter ideas already disfavored by other indirect evidence.

Maju said...

Oops, just deleted my comment because I realized that what I understood first was simply wrong after Mitchell's comment and a re-read of your entry (I did not see your reply then).

You have edited the original entry, right? Either that or I was quite dense when I read it first because I got exactly the wrong interpretation (>5 GeV/c² instead of <5 GeV/c²). My apologies.

... "it should have a mass that is about 2000 times lighter than an electron and about 2000+ times heavier than neutrino"...

Any idea of what kind of particle could have that kind of weight? I can't think of any fundamental particle that fits the requirements, so, if anything, it should be some sort of e-neutrino aggregate (but such thing isn't known to exist, is it?)

andrew said...

The $64,000 question in particle physics is what would have that weight.

No known particles are anywhere close. A composite particles from mesons to baryons to glueballs to tetraquarks to pentaquarks to atoms to molecules are too heavy. All charged leptons (electrons, muons, tauons) are too heavy as would any hypothetical composite particle made of leptons. All neutrinos are too light. Gluons and photons and hypothetical gravitons have zero rest mass (and the dynamic mass of gluons is far too great), and W and Z bosons and the Higgs boson are far too heavy.

SUSY superpartners (apart from the gravitino) are ruled out for masses of less than 5 GeV by and large by the LHC (SUSY purists would disagree).

Mitchell rightly notes that a right handed neutrino might, or some larger class of "sterile neutrinos" unrelated to any existing particle in the existing Standard Model particle zoo.

Axions, generally speaking are too light (anything up to about 10 eV would look like an additional neutrino type in cosmic background radiation, e.g., and would obliterate galaxy level structure in the universe). Neutrino condensates have been advanced but not very convincingly.

Pretty much any dark matter candidate has to come "out of nowhere".