A few days ago i heard one professor enthusiastically talk about the new results from his group, one of many searching for dark matter. Another professor (I don't know him or what field he is working in) asked him why anyone would be interested in yet another experiment reporting that they have seen no dark matter. Good question. It's really a strange thing, to work year after year with many different methods, and be happy with zero results.
Of course, as people are quick to point out, a zero result is still a result. You know something more about the thing you are hunting for, you can rule out theoretical models and narrow down the search. This is of course why the results are reported and why they are closely watched by the theorists as well as experimentalists.
As I have stated before, it is actually the challenge of searching for nearly undetectable things that attracted me to particle physics, then neutrino detection, and then to dark matter searches. There are all sorts of marvelous efforts going on, to think of ways to detect a dark matter particle (we usually talk about WIMPs, for Weakly Interacting Massive Particles) and distinguish it from other particles that are more prone to interact with matter and give signals in a detector.
One thing that is exploited in dark matter searches is the fact that most forms of radiation (i.e. matter particles or photons) interact with the electrons in matter and don't do much to the atomic nuclei. Atoms are like fluffy clouds of electrons, with a tiny nucleus hiding inside (you might remember the Rutherford experiment). It is way more likely for anything that can interact which the electric fields in an atom to encounter one or more electrons, and perhaps knock them out of the atom, than to slam into the nucleus.
A WIMP ignores the electrons. It is blind to the electric field since it has no electric charge of its own, and it's extremely near-sighted. The weak interaction has a short range, so the probability that the WIMP will actually notice and interact with a nucleus is low -- but it is not zero. If it interacts, it will give some of it's energy to the nucleus, which will recoil.
When a recoiling nucleus moves through the matter in the detector it will also knock off electrons from the atoms it passes through, but since it is heavy and typically carries a lot of electric charge it will do it more efficiently than other particles. If you can build a detector that can identify this pattern and separate it from other interactions, you will already have reduced most of the things that can obscure the visibility of the very rare events where a WIMP actually interacts. This is the basic idea behind many of the efforts to actually find WIMPs.
One example is the CDMS detector. The idea behind this is to use two types of signals. They look at the total ionization, which is the number of electrons knocked out from their atoms. They also look at vibrations that occur when a particle is removed from its place in a crystal lattice. (These vibrations are called phonons, since they can be thought of -- and mathematically treated -- as "sound particles".) By comparing the amount of electrons with the total phonon energy they can identify recoiling nuclei.
It gets a little more complicated, since there are other things that can make a nucleus recoil, for example neutrons. Neutrons, like other kinds of radiation, can come from natural radioactivity and can also be created when cosmic rays interact with matter. In an experiment like this you have to be very careful with the neutrons. Neutrons are absorbed in materials which contains very light elements, like water or plastic that contains lots of hydrogen, so these can be used for shielding. It also helps to put the detectors deep underground, where you are safe from most of the cosmic rays. You also have to take extreme care not to get radioactive elements inside the detector. Interesting challenges. But we are getting very good at it.
The CDMS collaboration has a preprint out (it's one of the first links on that web page if you want to read the actual paper) with a limit that was first presented at a conference last week. A limit, which means that they with a high degree of confidence can say that the WIMP cannot interact better than so much, or else they would have been able to see it. Little by little, we narrow down the possible properties of a dark matter particle.
When I'm talking about CDMS I just have to say that their educational pages have a very high level and don't seem to be very accessible for people without physics background. Consider this sentence: "The CDMS experiments (and many others) aim to measure the recoil energy imparted to detector nuclei through neutralino-nucleon collisions by employing sensitive phonon detection equipment coupled to arrays of cryogenic germanium and silicon crystals." Or just the general heavy prose of this one: "The supersymmetric standard model (SUSY) offers a promising framework for expectations of particle species which could satisfy the observed properties of dark matter." That is clearly not the kind of language that works for outreach -- too many difficult words.
As an indication: the reading level of their direct detection page where I took those quotes is "Genius" (while that of this blog is "High school", or at least was before I quoted those things).
Now please tell me if what I have written in this post makes sense!