≡ Menu

Breaking the Speed of Light

The Large Hadron Collider at CERN (CERN/LHC/GridPP)

It’s been a tenet of the standard model of physics for over a century. The speed of light is a unwavering and unbreakable barrier, at least by any form of matter and energy we know of. Nothing in our Universe can travel faster than 299,792 km/s (186,282 miles per second), not even – as the term implies – light itself. It’s the universal constant, the “c” in Einstein’s E = mc2, a cosmic speed limit that can’t be broken.

That is, until now.

An international team of scientists at the Gran Sasso research facility outside of Rome announced today that they have clocked neutrinos traveling faster than the speed of light. The neutrinos, subatomic particles with very little mass, were contained within beams emitted from CERN 730 km (500 miles) away in Switzerland. Over a period of three years, 15,000 neutrino beams were fired from CERN at special detectors located deep underground at Gran Sasso. Where light would have made the trip in 2.4 thousandths of a second, the neutrinos made it there 60 nanoseconds faster – that’s 60 billionths of a second – a tiny difference to us but a huge difference to particle physicists!

The implications of such a discovery are staggering, as it would effectively undermine Einstein’s theory of relativity and force a rewrite of the Standard Model of physics.

The OPERA Neutrino Detector. Credit: LGNS.

“We are shocked,” said project spokesman and University of Bern physicist Antonio Ereditato.

“We have high confidence in our results. We have checked and rechecked for anything that could have distorted our measurements but we found nothing. We now want colleagues to check them independently.”

Neutrinos are created naturally from the decay of radioactive materials and from reactions that occur inside stars. Neutrinos are constantly zipping through space and can pass through solid material easily with little discernible effect… as you’ve been reading this billions of neutrinos have already passed through you!

The experiment, called OPERA (Oscillation Project with Emulsion-tRacking Apparatus) is located in Italy’s Gran Sasso facility 1,400 meters (4,593 feet) underground and uses a complex array of electronics and photographic plates to detect the particle beams. Its subterranean location helps prevent experiment contamination from other sources of radiation, such as cosmic rays. Over 750 scientists from 22 countries around the world work there.

Ereditato is confident in the results as they have been consistently measured in over 16,000 events over the past two years. Still, other experiments are being planned elsewhere in an attempt to confirm these remarkable findings. If they are confirmed, we may be looking at a literal breakdown of the modern rules of physics as we know them!

“We have high confidence in our results,” said Ereditato. “We have checked and rechecked for anything that could have distorted our measurements but we found nothing. We now want colleagues to check them independently.”

A preprint of the OPERA results will be posted on the physics website ArXiv.org.

Read more on the Nature article here and on Reuters.com.

UPDATE: The OPERA team paper can be found here.




A graphic designer in Rhode Island, Jason writes about space exploration on his blog Lights In The Dark, Discovery News, and, of course, here on Universe Today. Ad astra!

Comments on this entry are closed.

  • Anonymous September 23, 2011, 4:47 PM

    Free of Einstein’s cage? I’ve always hoped so, but then there’s the problem of colliding with 10^6 Hydrogen atoms per cubic metre at > c, we’re going to need a shield. I love TL’s idea of failing to account for Earth’s curvature though, that would that be embarrassing. All in all I’m glad I was born when I was, everything seems to be turning up surprises right now. The advanced supernova neutrino burst is an interesting thought but if neutrinos can exceed c then why does it have to be by a fixed amount?

  • Anonymous September 23, 2011, 5:44 PM

    And if the results prove to be accurate and some things actually can travel faster than the speed of light, so what? I always thought that is what science is all about, learning new things using the scientific method. Is it possible to learn new things without occasionally learning that previous learning was incorrect? I don’t think so. We ought to always keep our minds open.

  • Torbjörn Larsson September 23, 2011, 6:08 PM

    “We don’t allow FTL neutrinos here”, said the barman. A neutrino walks into a bar.

    [HT Miscience]

  • Bob Sireno September 23, 2011, 8:13 PM

    No. The photons travel through space at the same speed. If you were traveling at nearly the speed of light the photons would slowly leave your flashlight. Space is not analog.

    • Anonymous September 23, 2011, 10:12 PM

      Indeed! But, space is actually not empty.

  • Anonymous September 24, 2011, 12:08 AM

    I formulated the existence de speeds of some particles in the vacuum greater than c in my paper “VELOCIDADES MAYORES QUE LA VELOCIDAD DE LA LUZ” (“Speeds greater than the speed of the light”) published in 1969 in the newspaper “El Siglo” in Colombia. The present discovery on the speed of the neutrinos is the beginning of the show of the speeds superluminaries.

  • Anonymous September 24, 2011, 12:14 AM

    I formulated my theory on the existence de speeds of some particles in the vacuum greater than c in my paper “Velocidades mayores que la velocidad de la luz” (“Speeds greater than the speed of the light”) published in 1969 in the newspaper “El Siglo” in Colombia. The present discovery on the speed of the neutrinos is the beginning of the show of the speeds superluminaries.

  • Johnny White September 24, 2011, 12:19 AM

    If they are right, one day we may say: “If I were made of neutrinos…” instead of “If I could turn back time…”

  • Torbjörn Larsson September 24, 2011, 11:42 AM

    Of course, for example here is Phillip Gibbs, an “independent physicist” (i.e. most likely a general crackpot; at the very least a facilitator of such through his ViXra clone of Arxiv), who nevertheless makes some valuable comments:

    “The only physical idea that would correspond to a lack of energy dependence [seen in the CERN results] would be if the universe had two separate fixed speeds, one for neutrinos and one for photons. I don’t think such a theory could be made to work, and even if it did you would have to explain why the SN1987A neutrinos were not affected. I think the conclusion has to be that there is no new physical effect, just a systematic error that the collaboration needs to find.”

  • Anonymous September 24, 2011, 12:22 PM

    If you listen to the media, and don’t have a science education, then you might think there was a serious chance that scientists have broken the speed of light barrier. Here’s my digested version of the recent FTL hoopla…

    * Einstein was wrong?

    Not yet, he isn’t. Science is all connected. It would be difficult to break the speed of light limit and not break the theories that allow your GPS or your computer to work. Similar neutrinos from distant astronomical events arrive after the light. The paper cites the SN1987A supernova evidence.

    The paper’s authors are saying that the measured time of flight of the neutrinos is out by 60 nanoseconds: if this measurement were accurate then the neutrinos were going faster than light. They practically say that result is so insane, there has almost has to be something systematically wrong with the measurements. For completeness, we include the possibility of FTL particles, but we don’t really believe it ourselves.

    * The timing is out by 60 nanoseconds

    60 nanoseconds is 120 cycles of a 2 GHz clock, like the one you may have in your computer. It is 2 meters at the speed of light. This is not a tiny difference to a huge sum like the Pioneer anomaly.

    We can synchronize clocks to much better than this. We do not need to race light alongside the neutrinos. If we have a fibre optic link, we can send a pulse from CERN CNGS to OPERA, and back again. If you know the time for the round-trip, then half the time is the time taken to go one way, with maybe a tiny correction if one end is higher than the other, and the light is doing work against gravity.

    We can measure 700 Km to much better than 2 meters. There is a lovely graph showing the distance plotted against time, showing the shifts in 3D. The continental drift of 1 cm/year shows up clearly. You can see a sudden jump of 7 cm that coincided with the L’Aquila earthquake.

    * Is the result repeatable?

    The authors quote a six sigma limit on the result. The ‘sigma’ or standard deviation is a measure of how far away from the average value you are on a standard ‘bell’ cure graph. Six sigma means that the odds of this occurring naturally from the known random variables is 3.4 in a million.

    This shows that they have done about as much as is sensible to eliminate random error. However, if they are measuring from the wrong place, they will have exactly the same error every time. This is not a random error but a systematic error. I could get the same probabilities by averaging thousands of measurements with an inaccurate ruler, but that would not make me right.

    * Okay, Mister Wise-Ass, what is it then?

    Well, we don’t know. The OPERA detector is a big thing, and the CERN CNGS source is far from small. Remember that the Hubble telescope was originally assembled to an accuracy of 1/400th of a wavelength of light, tested, sent into space, and then found to be half an inch out. It would be easy to lose a meter or so trying to work out where exactly the detector plane lies within OPERA, or something like that.

    It may not be a spatial error: the timing is pretty complicated too. The detector does not instantaneously detect neutrinos, so you have to correct for all the delays in the system, and maybe one of those delays has been overdone.

    * So, how can we know what’s going on?

    The experimenters are doing the right thing. They looked for an obvious solution. They could not spot it, so they are opening the problem up for others to look at it. Sometimes, when you have worked on the same apparatus for years, a newcomer is able to spot something that you have missed.

    They have already repeated the experiment on the same apparatus a lot of times.
    It would be nice to reproduce this in a separate lab, but there isn’t really another facility quite like this just now. They aren’t cheap. If we really cannot sort this problem, we are going to have to build something.

    If we made another OPERA at half the distance, we would get a timing difference of 60 nanoseconds if there is a timing error, or 30 nanoseconds if the particles were really going faster than light. This would be expensive, but it might be practial to make another smaller detector much closer to CERN CNGS. If you are 30 times closer then you ought to get 30×30 or about a thousand times the signal. If we are 30 times closer and we still get the 60 nanoseconds timing error, then we know something is wrong.

    My guess is the mistake will be found, and we won’t have to build an extra detector. If we build an extra detector, the most likely result is still that we find a systematic error. If we build a second detector and the speed still comes out the same fraction faster than light, then we may start talking about faster than light particles with some small measure of belief that they may really exist.

    The truth is out there somewhere. It’s probably rather dull.

    • Torbjörn Larsson September 24, 2011, 4:28 PM

      FWIW, in 60 ns light travel ~ 18 m.

      It is also reasonable, since a 2 GHz chip would be a few millimeters long, so 120 clock cycles traveling the chip through transistor gates would be a factor 10 or more slower.

      If you get down to compare with 2 meters light travel, you have shrunk the chips and/or quickened the transistors *a lot*.

      • Anonymous September 25, 2011, 1:00 PM

        Yep. Damn. I wrote that rather too fast, and under the influence of a BBC news mention of “evidence for faster than light particles”, and dropped a power of ten. Bah.

        18 meters is a big error for a survey, so this would tend to point towards a timing error. However, the paper has a lot of authors, and I expect they have thought along these lines before they posted.


  • Anonymous September 25, 2011, 5:06 AM

    Buddha discovered 2554 years ago.

  • Anonymous September 25, 2011, 5:09 AM

    Buddha discovered 2554 years ago.

  • Anonymous September 26, 2011, 9:13 AM

    “But if something can travel some particular speed, something else must be able to go faster all the way up to and including infinity http://bit.ly/qsvnNe

  • Anonymous October 3, 2011, 10:19 PM

    See full letter at: http://www.3d-computing.com/pb/bbc_news_breaking_speed-of-light.pdf

    Dear Antonio Ereditato,

    I am taking you up on your appeal for scrutiny of the results from the OPERA experiment that you made public at a BBC News interview on September 22nd, 2011 (http://www.bbc.co.uk/news/science-environment-15017484) :

    “we want just to be helped by the community in understanding our crazy result – because it is crazy”.

    When I heard your name on TV, in newspapers, etc. I was reminded of, and hopefully you will recall, the several dinners and gatherings over several years we shared at CERN 30 years ago with Luciano Ramello, Tiziano Camporesi, Mario Caria, Vittorio Remondino, etc. when I was working at CERN on the trigger of the Delphi experiment.

    Here is my scrutiny of the results of your experiment.

    I will first summarize what you stated – that you measured the time it took for a neutrino beam to travel a distance of 732 Km, and it took 60 nanoseconds less time than it would have taken a light beam to travel the same distance.

    After trying and failing to find any errors you stated:

    “We wanted to find a mistake –trivial mistakes, more complicated mistakes, or nasty effects – and we didn’t. When you don’t find anything, then you say ‘well, now I’m forced to go out and ask the community to scrutinise this‘. ”

    The most logical step to have taken the first time you found this “crazy” result three years ago would have been to plan a specific low cost experiment that would clarify this discrepancy. Instead, you kept doing and redoing the same experiment for three years acquiring “travel times of neutrino bunches some 16,000 times”! Remember Albert Einstein’s definition of insanity: Doing the same thing over and over again and expecting different results.”

    You needed to clarify the discrepancy of a Time of Flight (TOF) measurement of 60 nanoseconds (equivalent to 60,000 picoseconds) over a total duration of 2.43 milliseconds.

    There exist thousands of articles describing apparatus (detectors and electronics) that can make accurate TOF measurement with a resolution as precise as 10 picoseconds (or even 6 picoseconds as presented some time ago at a workshop at Stanford Linear Accelerator –SLAC-).

    So this is what I would have done three years ago:

    Since the discrepancy you found is so big, you do not need the neutrino to travel 732 Km in order to see this discrepancy. Using simple proportions traveling just 3.2 Km distance should give you a difference of about 260 picoseconds which is 26 times greater than the 10 picoseconds resolution step of your measuring apparatus.

    Remember Galileo’s simple experiment of taking two stones of different weights and dropping them at the same instant from the leaning tower of Pisa and then verifying that they reached the ground at the same time.

    Now, send two beams in burst of bunches (these can be sent separately, but simultaneously and repetitively will facilitate viewing the minimum differences in real-time), one light and one neutrino in the underground 3.2 Km SLAC tunnel (or take two LHC experiment sites A and B at CERN of approximately 5 to 7 Km distance from each other, send neutrino bunches underground and send light bunches between 2 towers above ground from site A to site B); build two identical electronic channel circuits (or purchase off-the-shelf components) with a time resolution of 10 picoseconds (or 6 picoseconds) to measure the traveling time of the two beams; within the detector, use the same transducer (if possible) to convert light into electrical signals (e.g. Photek PMT240 or fast SiPM from Hamamatsu or STMicroelectronics); then SWAP THE TWO ELECTRONIC CHANNEL CIRCUITS and repeat the experiment on the two burst of bunches to make sure that the neutrino beam (as well as the light beam) has the same speed regardless of which electronic channel is used (otherwise the fault would lie in the measuring device). By synchronizing the start of the burst of bunches, at the arrival point you would be able to see in real time with a fast oscilloscope the minimum difference on travel time between the two beams (just like two cars racing).

    If the discrepancy (neutrino bunches arriving faster than the light bunches) still persists, then you should be confident of having done a diligent accurate experiment and others should be able to obtain the same results.

    The reason for measuring the speed of the light beam as well as the neutrino beam is not because it is necessary to measure one more time the speed of light which has been measured thousands of times, but the goal is to test the accuracy or your electronics (achieved by swapping electronic channel circuits), with the speed of light as your source of calibration or reference. Believe me, it was not trivial for me to design DUT (Device Under Test) boards at the Superconducting Super Collider in 1992 for the HP8200 half-million dollar test station of integrated circuits with a time resolution accuracy of 50 picoseconds at all pins of the device under test. I was certain that my design and circuit implementation were correct only by comparing signals.

    In scientific research it is necessary to master calculations that will predict specific results. It is necessary to master how to conduct an experiment, to master the knowledge of electronics, detectors, and all components that will be used in the experiment. It is necessary to know the expected results, reproducible by different instrumentation (swapping electronic channel circuits), and ultimately confirm or reject calculations. Only then is money not wasted, experiments can confirm calculations, and the door to progress is opened. Here it looks like nothing has been mastered because the discrepancy is not explained with calculations, it is not explained from the results of the experiment, and it is just called “crazy results.” Results from untested performance of the measuring instrumentation (achievable by swapping electronic channel circuits that measure two parameters), risks alarming many people, discredits scientific research and wastes a lot of newspaper ink and TV time.

    Have you calculated how much money you and your collaborators have spent these past three years on conducting inconclusive tests instead of comparing in a scientific way the speed of two beams?

    There would be no need to mention the word “crazy” after having conducted this scientific procedure.

    In research it is important to discuss, identify, and implement scientific procedures that allow the laws of nature to be understood. A dialogue is key to identifying the most cost effective scientific procedure that will yield the most accurate results.

    I hope this example of a scientific procedure to check if neutrino breaks the speed of light helps.

    Best regards,

    Dario Crosetto
    900 Hideaway Pl.
    DeSoto, TX, 75115 – USA

    Email: crosetto@att.net or info@3d-computing.com

    Link to the BBC NEWS: “Speed-of-light results under scrutiny at Cern”