Weakly interacting massive particles
Weakly interacting massive particles (WIMPs) are hypothetical particles that are thought to constitute dark matter. There exists no clear definition of a WIMP, but broadly a WIMP is a new elementary particle which interacts via gravity and any other force (or forces), potentially not part of the standard model itself, which is as weak as or weaker than the weak nuclear force, but also non-vanishing in its strength. A WIMP must also have been produced thermally in the early Universe, similarly to the particles of the standard model according to Big Bang cosmology, and usually will constitute cold dark matter. Obtaining the correct abundance of dark matter today via thermal production requires a self-annihilation cross section of , which is roughly what is expected for a new particle in the 100 GeV mass range that interacts via the electroweak force. Because supersymmetric extensions of the standard model of particle physics readily predict a new particle with these properties, this apparent coincidence is known as the “WIMP miracle”, and a stable supersymmetric partner has long been a prime WIMP candidate. However, recent null results from direct detection experiments along with the failure to produce evidence of supersymmetry in the Large Hadron Collider (LHC) experiment has cast doubt on the simplest WIMP hypothesis. Experimental efforts to detect WIMPs include the search for products of WIMP annihilation, including gamma rays, neutrinos and cosmic rays in nearby galaxies and galaxy clusters; direct detection experiments designed to measure the collision of WIMPs with nuclei in the laboratory, as well as attempts to directly produce WIMPs in colliders, such as the LHC.
Theoretical framework and properties
|This section needs additional citations for verification. (January 2011) (Learn how and when to remove this template message)|
WIMP-like particles are predicted by R-parity-conserving supersymmetry, a popular type of extension to the standard model of particle physics, although none of the large number of new particles in supersymmetry have been observed. WIMP-like particles are also predicted by universal extra dimension and little Higgs theories.
|SUSY||R-parity||lightest supersymmetric particle (LSP)|
|UED||KK-parity||lightest Kaluza-Klein particle (LKP)|
|little Higgs||T-parity||lightest T-odd particle (LTP)|
The main theoretical characteristics of a WIMP are:
- Interactions only through the weak nuclear force and gravity, or possibly other interactions with cross-sections no higher than the weak scale;
- Large mass compared to standard particles (WIMPs with sub-GeV masses may be considered to be light dark matter).
Because of their lack of electromagnetic interaction with normal matter, WIMPs would be dark and invisible through normal electromagnetic observations. Because of their large mass, they would be relatively slow moving and therefore "cold". Their relatively low velocities would be insufficient to overcome the mutual gravitational attraction, and as a result, WIMPs would tend to clump together. WIMPs are considered one of the main candidates for cold dark matter, the others being massive compact halo objects (MACHOs) and axions. (These names were deliberately chosen for contrast, with MACHOs named later than WIMPs.) Also, in contrast to MACHOs, there are no known stable particles within the standard model of particle physics that have all the properties of WIMPs. The particles that have little interaction with normal matter, such as neutrinos, are all very light, and hence would be fast moving, or "hot".
WIMPs as dark matter
Although the existence of WIMPs in nature is hypothetical at this point, it would resolve a number of astrophysical and cosmological problems related to dark matter. There is near consensus today among astronomers that most of the mass in the Universe is dark. Simulations of a universe full of cold dark matter produce galaxy distributions that are roughly similar to what is observed. By contrast, hot dark matter would smear out the large-scale structure of galaxies and thus is not considered a viable cosmological model.
The WIMP fits the model of a relic dark matter particle from the early Universe, when all particles were in a state of thermal equilibrium. For sufficiently high temperatures, such as those existing in the early Universe, the dark matter particle and its antiparticle would have been both forming from and annihilating into lighter particles. As the Universe expanded and cooled, the average thermal energy of these lighter particles decreased and eventually became insufficient to form a dark matter particle-antiparticle pair. The annihilation of the dark matter particle-antiparticle pairs, however, would have continued, and the number density of dark matter particles would have begun to decrease exponentially. Eventually, however, the number density would become so low that the dark matter particle and antiparticle interaction would cease, and the number of dark matter particles would remain (roughly) constant as the Universe continued to expand. Particles with a larger interaction cross section would continue to annihilate for a longer period of time, and thus would have a smaller number density when the annihilation interaction ceases. Based on the current estimated abundance of dark matter in the Universe, if the dark matter particle is such a relic particle, the interaction cross section governing the particle-antiparticle annihilation can be no larger than the cross section for the weak interaction. If this model is correct, the dark matter particle would have the properties of the WIMP.
Because WIMPs may only interact through gravitational and weak forces, they are extremely difficult to detect. However, there are many experiments underway to attempt to detect WIMPs both directly and indirectly. Indirect detection refers to the observation of annihilation or decay products of WIMPs far away from Earth. Indirect detection efforts typically focus on locations where WIMP dark matter is thought to accumulate the most: in the centers of galaxies and galaxy clusters, as well as in the smaller satellite galaxies of the Milky Way. These are particularly useful since they tend to contain very little baryonic matter, reducing the expected background from standard astrophysical processes. Typical indirect searches look for excess gamma rays, which are predicted both as final-state products of annihilation, or are produced as charged particles interact with ambient radiation via inverse Compton scattering. The spectrum and intensity of a gamma ray signal depends on the annihilation products, and must be computed on a model-by-model basis. Experiments that have placed bounds on WIMP annihilation, via the non-observation of an annihilation signal, include the Fermi-LAT gamma ray telescope and the VERITAS ground-based gamma ray observatory. Although the annihilation of WIMPs into standard model particles also predicts the production of high-energy neutrinos, their interaction rate is too low to reliably detect a dark matter signal at present. Future observations from the IceCube observatory in Antarctica may be able to differentiate WIMP-produced neutrinos from standard astrophysical neutrinos; however, by 2014, only 37 cosmological neutrinos had been observed, making such a distinction impossible.
Another type of indirect WIMP signal could come from the Sun. Halo WIMPs may, as they pass through the Sun, interact with solar protons, helium nuclei as well as heavier elements. If a WIMP loses enough energy in such an interaction to fall below the local escape velocity, it would not have enough energy to escape the gravitational pull of the Sun and would remain gravitationally bound. As more and more WIMPs thermalize inside the Sun, they begin to annihilate with each other, forming a variety of particles, including high-energy neutrinos. These neutrinos may then travel to the Earth to be detected in one of the many neutrino telescopes, such as the Super-Kamiokande detector in Japan. The number of neutrino events detected per day at these detectors depends on the properties of the WIMP, as well as on the mass of the Higgs boson. Similar experiments are underway to detect neutrinos from WIMP annihilations within the Earth and from within the galactic center.
Direct detection refers to the observation of the effects of a WIMP-nucleus collision as the dark matter passes through a detector in an Earth laboratory. While most WIMP models indicate that a large enough number of WIMPs must be captured in large celestial bodies for indirect detection experiments to succeed, it remains possible that these models are either incorrect or only explain part of the dark matter phenomenon. Thus, even with the multiple experiments dedicated to providing indirect evidence for the existence of cold dark matter, direct detection measurements are also necessary to solidify the theory of WIMPs.
Although most WIMPs encountering the Sun or the Earth are expected to pass through without any effect, it is hoped that a large number of dark matter WIMPs crossing a sufficiently large detector will interact often enough to be seen—at least a few events per year. The general strategy of current attempts to detect WIMPs is to find very sensitive systems that can be scaled up to large volumes. This follows the lessons learned from the history of the discovery and (by now) routine detection of the neutrino.
Cryogenic crystal detectors - A technique used by the Cryogenic Dark Matter Search (CDMS) detector at the Soudan Mine relies on multiple very cold germanium and silicon crystals. The crystals (each about the size of a hockey puck) are cooled to about 50 mK. A layer of metal (aluminium and tungsten) at the surfaces is used to detect a WIMP passing through the crystal. This design hopes to detect vibrations in the crystal matrix generated by an atom being "kicked" by a WIMP. The tungsten transition edge sensors (TES) are held at the critical temperature so they are in the superconducting state. Large crystal vibrations will generate heat in the metal and are detectable because of a change in resistance. CRESST, CoGeNT, and EDELWEISS run similar setups.
Noble gas scintillators - Another way of detecting atoms "knocked about" by a WIMP is to use scintillating material, so that light pulses are generated by the moving atom and detected, often with PMTs. Experiments such as DEAP at SNOLAB and DarkSide at the LNGS instrument a very large target mass of liquid argon for sensitive WIMP searches. ZEPLIN, and XENON used xenon to exclude WIMPs at higher sensitivity, with the most stringent limits to date provided by the XENON1T detector, utilizing 3.5 tons of liquid xenon. Even larger multi-ton liquid xenon detectors have been approved for construction from the XENON, LUX-ZEPLIN and PandaX collaborations.
Crystal scintillators - Instead of a liquid noble gas, an in principle simpler approach is the use of a scintillating crystal such as NaI(Tl). This approach is taken by DAMA/LIBRA, an experiment that observed an annular modulation of the signal consistent with WIMP detection (see #Recent Limits). Several experiments are attempting to replicate those results, including ANAIS and DM-Ice, which is codeploying NaI crystals with the IceCube detector at the South Pole. KIMS is approaching the same problem using CsI(Tl) as a scintillator.
Bubble chambers - The PICASSO (Project In Canada to Search for Supersymmetric Objects) experiment is a direct dark matter search experiment that is located at SNOLAB in Canada. It uses bubble detectors with Freon as the active mass. PICASSO is predominantly sensitive to spin-dependent interactions of WIMPs with the fluorine atoms in the Freon. COUPP, a similar experiment using trifluoroiodomethane(CF3I), published limits for mass above 20 GeV in 2011.
A bubble detector is a radiation sensitive device that uses small droplets of superheated liquid that are suspended in a gel matrix. It uses the principle of a bubble chamber but, since only the small droplets can undergo a phase transition at a time, the detector can stay active for much longer periods.[clarification needed] When enough energy is deposited in a droplet by ionizing radiation, the superheated droplet becomes a gas bubble. The bubble development is accompanied by an acoustic shock wave that is picked up by piezo-electric sensors. The main advantage of the bubble detector technique is that the detector is almost insensitive to background radiation. The detector sensitivity can be adjusted by changing the temperature, typically operated between 15 °C and 55 °C. There is another similar experiment using this technique in Europe called SIMPLE.
PICASSO reports results (November 2009) for spin-dependent WIMP interactions on 19F, for masses of 24 Gev new stringent limits have been obtained on the spin-dependent cross section of 13.9 pb (90% CL). The obtained limits restrict recent interpretations of the DAMA/LIBRA annual modulation effect in terms of spin dependent interactions.
PICO is an expansion of the concept planned in 2015.
Other types of detector - Time projection chambers (TPCs) filled with low pressure gases are being studied for WIMP detection. The Directional Recoil Identification From Tracks (DRIFT) collaboration is attempting to utilize the predicted directionality of the WIMP signal. DRIFT uses a carbon disulfide target, that allows WIMP recoils to travel several millimetres, leaving a track of charged particles. This charged track is drifted to an MWPC readout plane that allows it to be reconstructed in three dimensions and determine the origin direction. DMTPC is a similar experiment with CF4 gas.
There are currently no confirmed detections of dark matter from direct detection experiments, with the strongest exclusion limits coming from the LUX and SuperCDMS experiments, as shown in figure 2. With 370 kilograms of xenon LUX is more sensitive than XENON or CDMS. First results from October 2013 report that no signals were seen, appearing to refute results obtained from less sensitive instruments. and this was confirmed after the final data run ended in May 2016.
Historically there have been four anomalous sets of data from different direct detection experiments, two of which have now been explained with backgrounds (CoGeNT and CRESST-II), and two which remain unexplained (DAMA/LIBRA and CDMS-Si). In February 2010, researchers at CDMS announced that they had observed two events that may have been caused by WIMP-nucleus collisions.
CoGeNT, a smaller detector using a single germanium puck, designed to sense WIMPs with smaller masses, reported hundreds of detection events in 56 days. They observed an annual modulation in the event rate that could indicate light dark matter. However a dark matter origin for the CoGeNT events has been refuted by more recent analyses, in favour of an explanation in terms of a background from surface events.
Annual modulation is one of the predicted signatures of a WIMP signal, and on this basis the DAMA collaboration has claimed a positive detection. Other groups, however, have not confirmed this result. The CDMS data made public in May 2004 exclude the entire DAMA signal region given certain standard assumptions about the properties of the WIMPs and the dark matter halo, and this has been followed by many other experiments (see Fig 2, right).
The future of direct detection
The next decade should see the emergence of several multi-tonne mass direct detection experiments, which will probe WIMP-nucleus cross sections orders or magnitude smaller than the current state-of-the-art sensitivity. Examples of such next-generation experiments are LUX-ZEPLIN (LZ), which will start as a several tonne mass liquid xenon experiment before moving up to twenty tonnes, and DARWIN, another proposed liquid xenon direct detection experiment which will have a target mass approaching twenty tonnes.
Such multi-tonne experiments will also face a new background in the form of neutrinos, which will limit their ability to probe the WIMP parameter space beyond a certain point, known as the neutrino floor. However although its name may imply a hard limit, the neutrino floor represents the region of parameter space beyond which experimental sensitivity can only improve at best as the inverse square root of exposure (the product of detector mass and running time). For WIMP masses below 10 GeV the dominant source of neutrino background is from the sun, while for higher masses the background contains contributions from atmospheric neutrinos and the diffuse supernova neutrino background.
- Jungman, Gerard; Kamionkowski, Marc; Griest, Kim (1996). "Supersymmetric dark matter". Physics Reports. 267 (5–6): 195. Bibcode:1996PhR...267..195J. arXiv: . doi:10.1016/0370-1573(95)00058-5.
- LHC discovery maims supersymmetry again, Discovery News
- Craig, Nathaniel (2013). "The State of Supersymmetry after Run I of the LHC". arXiv: [hep-ph].
- Fox, Patrick J.; Jung, Gabriel; Sorensen, Peter; Weiner, Neal (2013). "Dark Matter in Light of LUX". Physical Review D. 89 (10): 103526. Bibcode:2014PhRvD..89j3526F. arXiv: . doi:10.1103/PhysRevD.89.103526.
- Klapdor-Kleingrothaus, H. V. (1998). "Double Beta Decay and Dark Matter Search - Window to New Physics now, and in future (GENIUS)". In V. Klapdor-Kleingrothaus, H. Paes. Beyond the Desert. 1997. IOP. p. 485. Bibcode:1998hep.ex....2007K. arXiv: .
- Kamionkowski, Marc (1997). "WIMP and Axion Dark Matter". arXiv: .
- Zacek, Viktor (2007). "Dark Matter". Fundamental Interactions. arXiv: [astro-ph]. doi:10.1142/9789812776105_0007.
- Griest, Kim (1993). "The Search for the Dark Matter: WIMPs and MACHOs". Annals of the New York Academy of Sciences. 688: 390–407. Bibcode:1993NYASA.688..390G. PMID 26469437. arXiv: . doi:10.1111/j.1749-6632.1993.tb43912.x.
- Griest, Kim (1991). "Galactic Microlensing as a Method of Detecting Massive Compact Halo Objects". The Astrophysical Journal. 366: 412–421. Bibcode:1991ApJ...366..412G. doi:10.1086/169575.
- Conroy, Charlie; Wechsler, Risa H.; Kravtsov, Andrey V. (2005). "Modeling Luminosity-Dependent Galaxy Clustering Through Cosmic Time". The Astrophysical Journal. 647: 201–214. Bibcode:2006ApJ...647..201C. arXiv: . doi:10.1086/503602.
- The Millennium Simulation Project, Introduction: The Millennium Simulation The Millennium Run used more than 10 billion particles to trace the evolution of the matter distribution in a cubic region of the Universe over 2 billion light-years on a side.
- Ackermann, M.; et al. (The Fermi-LAT Collaboration) (2014). "Dark matter constraints from observations of 25 Milky Way satellite galaxies with the Fermi Large Area Telescope". Physical Review D. 89 (4): 042001. Bibcode:2014PhRvD..89d2001A. arXiv: . doi:10.1103/PhysRevD.89.042001.
- Grube, Jeffrey; for the VERITAS Collaboration (2012). "VERITAS Limits on Dark Matter Annihilation from Dwarf Galaxies". AIP Conference Proceedings. AIP Conference Proceedings. 1505: 689. Bibcode:2012AIPC.1505..689G. arXiv: . doi:10.1063/1.4772353.
- Aartsen, M. G.; et al. (IceCube Collaboration) (2014). "Observation of High-Energy Astrophysical Neutrinos in Three Years of IceCube Data". Physical Review Letters. 113 (10): 101101. Bibcode:2014PhRvL.113j1101A. PMID 25238345. arXiv: . doi:10.1103/PhysRevLett.113.101101.
- Ferrer, F.; Krauss, L. M.; Profumo, S. (2006). "Indirect detection of light neutralino dark matter in the next-to-minimal supersymmetric standard model". Physical Review D. 74 (11): 115007. Bibcode:2006PhRvD..74k5007F. arXiv: . doi:10.1103/PhysRevD.74.115007.
- Freese, Katherine (1986). "Can scalar neutrinos or massive Dirac neutrinos be the missing mass?". Physics Letters B. 167 (3): 295. Bibcode:1986PhLB..167..295F. doi:10.1016/0370-2693(86)90349-7.
- Merritt, D.; Bertone, G. (2005). "Dark Matter Dynamics and Indirect Detection". Modern Physics Letters A. 20 (14): 1021–1036. Bibcode:2005MPLA...20.1021B. arXiv: . doi:10.1142/S0217732305017391.
- Fornengo, Nicolao (2006). "Status and perspectives of indirect and direct dark matter searches". Advances in Space Research. 41 (12): 2010–2018. Bibcode:2008AdSpR..41.2010F. arXiv: . doi:10.1016/j.asr.2007.02.067.
- Aprile, E; et al. (2017). "First Dark Matter Search Results from the XENON1T Experiment". arXiv: .
- Behnke, E.; Behnke, J.; Brice, S. J.; Broemmelsiek, D.; Collar, J. I.; Cooper, P. S.; Crisler, M.; Dahl, C. E.; Fustin, D.; Hall, J.; Hinnefeld, J. H.; Hu, M.; Levine, I.; Ramberg, E.; Shepherd, T.; Sonnenschein, A.; Szydagis, M. (10 January 2011). "Improved Limits on Spin-Dependent WIMP-Proton Interactions from a Two Liter Bubble Chamber". Physical Review Letters. 106 (2): 021303. Bibcode:2011PhRvL.106b1303B. PMID 21405218. arXiv: . doi:10.1103/PhysRevLett.106.021303.
- Bubble Technology Industries
- Archambault, S.; Aubin, F.; Auger, M.; Behnke, E.; Beltran, B.; Clark, K.; Dai, X.; Davour, A.; et al. (PICASSO Collaboration) (2009). "Dark Matter Spin-Dependent Limits for WIMP Interactions on 19F by PICASSO". Physics Letters B. 682 (2): 185. Bibcode:2009PhLB..682..185A. arXiv: . doi:10.1016/j.physletb.2009.11.019.
- Cooley, J. (28 October 2014). "Overview of non-liquid noble direct detection dark matter experiments". Physics of the Dark Universe. 4: 92–97. Bibcode:2014PDU.....4...92C. arXiv: . doi:10.1016/j.dark.2014.10.005.
- "New Experiment Torpedoes Lightweight Dark Matter Particles". 30 October 2013. Retrieved 6 May 2014.
- "First Results from LUX, the World’s Most Sensitive Dark Matter Detector". Berkeley Lab News Center. 30 October 2013. Retrieved 6 May 2014.
- Dark matter search comes up empty. July 2016
- "Largest-ever dark-matter experiment poised to test popular theory". Nature News & Comment. Retrieved 15 January 2017.
- Davis, Jonathan H. (2015). "The Past and Future of Light Dark Matter Direct Detection" (PDF). Int.J.Mod.Phys. A. 30 (15): 1530038. Bibcode:2015IJMPA..3030038D. arXiv: . doi:10.1142/S0217751X15300380.
- "Key to the universe found on the Iron Range?". Retrieved December 18, 2009.
- CDMS Collaboration. "Results from the Final Exposure of the CDMS II Experiment" (PDF).. See also a non-technical summary: CDMS Collaboration. "Latest Results in the Search for Dark Matter" (PDF).
- The CDMS II Collaboration (2010). "Dark Matter Search Results from the CDMS II Experiment". Science. 327 (5973): 1619–21. Bibcode:2010Sci...327.1619C. PMID 20150446. arXiv: . doi:10.1126/science.1186112.
- Eric Hand (2010-02-26). "A CoGeNT result in the hunt for dark matter". Nature News.
- C. E. Aalseth; CoGeNT collaboration (2011). "Results from a Search for Light-Mass Dark Matter with a P-type Point Contact Germanium Detector". Physical Review Letters. 106 (13): 131301. Bibcode:2011PhRvL.106m1301A. PMID 21517370. arXiv: . doi:10.1103/PhysRevLett.106.131301.
- James Dacey (June 2011). "CoGeNT findings support dark-matter halo theory". physicsworld. Retrieved 5 May 2015.
- Davis, Jonathan H.; McCabe, Christopher; Boehm, Celine (2014). "Quantifying the evidence for Dark Matter in CoGeNT data". JCAP. 1408 (8): 014. Bibcode:2014JCAP...08..014D. arXiv: . doi:10.1088/1475-7516/2014/08/014.
- Drukier, Andrzej K.; Freese, Katherine; Spergel, David N. (15 June 1986). "Detecting cold dark-matter candidates". Physical Review D. 33 (12): 3495–3508. Bibcode:1986PhRvD..33.3495D. doi:10.1103/PhysRevD.33.3495.
- K. Freese, J. Frieman, and A. Gould, (1988). "Signal Modulation in Cold Dark Matter Detection". Physical Review D. 37 (12): 3388–3405. Bibcode:1988PhRvD..37.3388F. doi:10.1103/PhysRevD.37.3388.
- Malling, D. C.; Akerib, D. S.; Araujo, H. M.; Bai, X.; Bedikian, S.; Bernard, E.; Bernstein, A.; Bradley, A.; Cahn, S. B.; Carmona-Benitez, M. C.; Carr, D.; Chapman, J. J.; Clark, K.; Classen, T.; Coffey, T.; Curioni, A.; Currie, A.; Dazeley, S.; de Viveiros, L.; Dragowsky, M.; Druszkiewicz, E.; Faham, C. H.; Fiorucci, S.; Gaitskell, R. J.; Gibson, K. R.; Hall, C.; Hanhardt, M.; Holbrook, B.; Ihm, M.; et al. (2011). "After LUX: The LZ Program". arXiv: [astro-ph.IM].
- Baudis, Laura (2012). "DARWIN: dark matter WIMP search with noble liquids". J. Phys. Conf. Ser. 375: 012028. arXiv: . doi:10.1088/1742-6596/375/1/012028.
- Billard, J.; Strigari, L.; Figueroa-Feliciano, E. (2014). "Implication of neutrino backgrounds on the reach of next generation dark matter direct detection experiments". Phys. Rev. D. 89 (2): 023524. Bibcode:2014PhRvD..89b3524B. arXiv: . doi:10.1103/PhysRevD.89.023524.
- Davis, Jonathan H. (2015). "Dark Matter vs. Neutrinos: The effect of astrophysical uncertainties and timing information on the neutrino floor". JCAP. 1503 (3): 012. Bibcode:2015JCAP...03..012D. arXiv: . doi:10.1088/1475-7516/2015/03/012.
- Bertone, Gianfranco (2010). Particle Dark Matter: Observations, Models and Searches. Cambridge University Press. p. 762. ISBN 978-0-521-76368-4.
- Cerdeno, David G.; Green, Anne M. (2010). "Direct detection of WIMPs". Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms. 269 (4): 498. Bibcode:2011NIMPB.269..498L. arXiv: . doi:10.1016/j.nimb.2010.12.064.
- Davis, Jonathan H. (2015). "The Past and Future of Light Dark Matter Direct Detection". Int. J. Mod. Phys. A. 30 (15): 1530038. Bibcode:2015IJMPA..3030038D. arXiv: . doi:10.1142/S0217751X15300380.
- Del Nobile, Eugenio; Gelmini, Graciela B.; Gondolo, Paolo; Huh, Ji-Haeng (2015). "Update on the Halo-independent Comparison of Direct Dark Matter Detection Data". Physics Procedia. 61: 45–54. Bibcode:2015PhPro..61...45D. arXiv: . doi:10.1016/j.phpro.2014.12.009.
- Particle Data Group review article on WIMP search (S. Eidelman et al. (Particle Data Group), Phys. Lett. B 592, 1 (2004) Appendix : OMITTED FROM SUMMARY TABLE)
- Timothy J. Sumner, Experimental Searches for Dark Matter in Living Reviews in Relativity, Vol 5, 2002
- Portraits of darkness, New Scientist, August 31, 2013. Preview only.