Forgot your password?
typodupeerror
Supercomputing Encryption Security Science Technology

Simple Mod Turns Diodes Into Photon Counters 118

Posted by timothy
from the measure-once-vote-often dept.
KentuckyFC writes "The standard way to detect single photons is to use an avalanche photodiode in which a single photon can trigger an avalanche of current. These devices have an important drawback, however. They cannot distinguish the arrival of a single photon from the simultaneous arrival of two or more. But a team of physicists in the UK has found a simple mod that turns avalanche photodiodes into photon counters. They say that in the first instants after the avalanche forms, its current is proportional to the number of photons that have struck. All you have to do is measure it at this early stage. That's like turning a Fiat 500 into a Ferrari. Photon counting is one of the enabling technologies behind optical quantum computing. A number of schemes are known in which it is necessary to count the arrival of 0, 1 or 2 photons at specific detectors (abstract). With such a cheap detector now available (as well as decent photon guns), we could see dramatic progress in this field in the coming months."
This discussion has been archived. No new comments can be posted.

Simple Mod Turns Diodes Into Photon Counters

Comments Filter:
  • Down already? (Score:1, Offtopic)

    by DrLudicrous (607375)
    Instantaneously slashdotted site?
  • So In other words... (Score:4, Informative)

    by Penguinisto (415985) on Tuesday July 08, 2008 @10:24AM (#24099799) Journal

    ...they treat the earliest stages as if the diode were a unijunction transistor of sorts (using the photon reaction as if it were the gate 'voltage'), and check the waveform? Gah - I'll RTFA(bstract) @ work... :)

    /P

    • by slmdmd (769525)
      I think this will usher in a new era where logic will be built around multiple states as opposed to binary(current 1/0 state of diode). The implications are huge. If else programming will be cave man tool. To my understanding this will make us create unimaginably complex/intelligent machines. In other words equal to humans if not better. The dream of matrix will come true. Consider if 1 then.. if 2 if 3 if 4 then... wow
      • I wouldn't be so sure of it happening soon... some odd thoughts:

        The oscilloscopes required to read and reliably analyze such a waveform at the moment cost approximately more than most folks' houses (not counting probes, tips, calibration contracts, etc). We have a few here where I work, and they're treated like newborn children because of the pricetag.

        Even in a digital waveform, you've always had rise and falloff, and those could conceivably be read and acted on (though in only a few damned-small niche case

        • by TigerNut (718742)
          Analog programming has been done for decades, using analog computers. Analog computers were the tools for initial research into chaotic (strange attractor) behavior. In the analog programming world, you configure a differential equation appropriate to the problem you're trying to solve, and the system produces a solution.

          Trying to recast the problems we've been solving (or creating) with digital computers in analog computer terms is an interesting exercise; some problems have analogs (pun not really intende

      • I think this will usher in a new era where logic will be built around multiple states as opposed to binary(current 1/0 state of diode). The implications are huge. If else programming will be cave man tool. To my understanding this will make us create unimaginably complex/intelligent machines. In other words equal to humans if not better. The dream of matrix will come true. Consider if 1 then.. if 2 if 3 if 4 then... wow

        Gee. I had no idea the Matrix was based on the switch statement...

  • by UnknowingFool (672806) on Tuesday July 08, 2008 @10:26AM (#24099821)
    You kids and your fancy diodes. Back in my day we counted photon by hand. Some people used paper to record the counts. We called them amateurs. Now get off my lawn!
  • Cooled devices? (Score:1, Interesting)

    by EmagGeek (574360)

    Cheap, cooled devices...

    They don't go into a whole lot of detail on what they mean by "cooled devices." This smells an awful lot like "you have to spend a ton of money to save a tiny bit of money."

    If these things have to be cooled with some liquefied gas, then I suppose any cost savings on the material or approach would be quickly eaten up by energy used to cool them... just a hunch...

    • Re:Cooled devices? (Score:5, Informative)

      by ZombieWomble (893157) on Tuesday July 08, 2008 @10:34AM (#24099933)
      I think you misread the article. Specifically where it says:

      Various people, including Shields himself, have come up with complex, cooled devices that can count photons.

      It is the current generation of photon counting detectors which typically require high degrees of cooling (usually with LN2, as you suggest). Photodiodes of the type discussed in the article typically don't have such extreme cooling requirements under normal operation, so presumably that's what's so nice about this mod, as well.

      • by EmagGeek (574360)

        You're right, I misread it. That's what I get for letting my work distract me from my /. reading.

        Hah...

      • Erm.. It doesn't matter what you're using for photon counting. You're going to have to chill it to cut the noise [wikipedia.org] down. Of course, for some applications, the noise might be below an acceptable level.

        • The noise from "real" photons (that is, those directly produced by black body radiation) is really rather small - vanishingly so, at visible and higher frequencies for an emitter at room temperature. Cooling in detectors at such energies is primarily needed for electrical noise caused in the system by events at thermal energies, something which is significantly less of a problem in avalanche photodiodes.

          Obviously, if you want to measure photons at thermal energies, then cooling is certainly a requirement.

        • by digitrev (989335)
          You're close. The blackbody radiation is often the signal that's getting drowned out by all the noise. For example, we have to cool our bolometers down to 300 mK to get any meaningful data. And that's just for testing. When the detector finally gets launched, it's going to have a huge tank of liquid Helium-3 to keep it cold (via a complicated pump system).
    • Re:Cooled devices? (Score:5, Informative)

      by Robotbeat (461248) on Tuesday July 08, 2008 @10:46AM (#24100121) Journal

      There's really no way around cooling the sensor for photon counting, especially if you use near-infrared or lower. If the sensor itself is giving off black-body radiation of the type you're looking for, then it's pretty much worthless to try to count photons because the laws of thermodynamics and quantum mechanics conspire against you. I can imagine visible or UV photon counting with uncooled sensors, but certainly not far-infrared. These thermally-generated photons are what cause the "dark count rate" of a device, and cooling the device can help reduce the dark count rate. Here you are, from wikipedia:
      http://en.wikipedia.org/wiki/Single-Photon_Avalanche_Diode [wikipedia.org]

      • Re:Cooled devices? (Score:5, Interesting)

        by mapsjanhere (1130359) on Tuesday July 08, 2008 @11:06AM (#24100413)
        You can buy a single photon counter optimized at 680 nm that works at RT. Unfortunately, they are $5k a pop. This way of using avalanche diodes for counting enables a lot of new technologies.
        Years ago we played with detecting high energy particles in a grid of scintillating fibers, but for a high precision array you just couldn't afford the detectors. Now I guess I can revisit that if the technology pans out.
  • As the physics is _not_ understood well enough to support the outrageous security claims that are being made by these grant-hungry researchers.

    My guess is that at some time the possibility to evasdrop on these connections (and there are already some succcesses with that, despite the initial claims of impossibility), will render this technology useless. I hoper there is at least some useful other discoveries alsong the way...

    • Not Quite (Score:5, Insightful)

      by ruin20 (1242396) on Tuesday July 08, 2008 @11:24AM (#24100731)
      Although I agree with you on the aspect of quantum encryption, that's not the only application for this technology. Quantum computing [wikipedia.org] is a means to increase the maximum processor speed exponentially. [caltech.edu] (see the "The Potential and Power of Quantum Computing" for a good explanation on how)

      a lot of the applications for "security" actually is the defeat of cryptanalysis systems as these computers could crack keys in a reasonable amount of time. This would start to drive key length to very large values in order to keep data safe.

      Essentially the value in quantum computing is you can set up a logical relationship between all the qbits and then preform an operation on any number of them and they instantaneously effect the remaining qbits. This saves the computation time for preforming operations on all the other qbits. The question on making this feasible is can you make the read/write time for each of the qbits reasonable and the technology affordable to do so. This seems to be a huge step in the right direction for the latter.

      • by gweihir (88907)

        Well, I happen to know some of the people in quantum computing. First, it is not exponential, just a factor of n for most practical computing problems. Then there is the problem of scalability. Most people do not know it, but current CPUs are speed limited by interconnect, i.e. the signal lines of the chip. With quantum computing, the interconnect problem is dramatically worse. I would think that quantum computers will not reach the power of CPUs 20 years ago in our lifetimes, if ever. There may still be so

  • "They say that in the first instants after the avalanche forms, its current is proportional to the number of photons that have struck. All you have to do is measure it at this early stage."

    That's nothing, I used to turn avalanche photodiodes into photon counters in my T-16 back home.

  • by 140Mandak262Jamuna (970587) on Tuesday July 08, 2008 @10:35AM (#24099939) Journal
    I mean if multiple photons arrive at the same time at the detector should they be counted as a single vote or multiple votes? Whatever you say someone or the other would object and eventually it will be decided in the Supreme Court. Counting is quite weird in Florida.
    • by JamesP (688957) on Tuesday July 08, 2008 @11:20AM (#24100671)

      You see, it depends on the wavelength of the photon.

      If they are red they may be counted multiple times

      If they are blue, counting is more difficult, and sometimes several pass without being counted

    • Well, since all these photons are causing changes in the waveform all you really get are dimpled chads so none of them really get counted. Now if we could only get a photon of infinite energy...
  • by H+FTW (1264808) on Tuesday July 08, 2008 @10:39AM (#24099995)

    Something that hasn't been pointed out is how useful this will be in high energy physics. The basic way of measuring a lot of particles is to look for the photons emitted when they interact with materials.

    This should help reduce the cost of certain detectors. Especially for measuring neutrinos that can only be spotted by the cherenkov radiation they give off as they pass through massive detectors (look here http://en.wikipedia.org/wiki/Cherenkov_radiation [wikipedia.org])

    • Re: (Score:3, Informative)

      by locofungus (179280)

      Neutrinos are not charged. Therefore they don't give off Cherenkov radiation.

      I'm under the impression that it's the "reverse" of n -> p + e + v_e'; n + v_e -> p + e that is the usual reaction that is needed - probably the e then gives off Cherenkov radiation.

      Tim.

      • by H+FTW (1264808)

        your correct

        given enough time the v_e- will interact to give off an e- that causes the cherenkov radiation

    • by mako1138 (837520)

      Reading the paper, I'm not so sure this is useful for HEP. In HEP, you are not interested in the number of photons, but rather their energy.

    • by juancn (596002)
      Having better neutrino detectors could pave the way for neutrino-based communications. This would be great for long distance communications (think Mars for example) given that neutrinos can travel very long distances without hitting anything. At least it would be fun to look for E.T. communications.
  • by FlyingSquidStudios (1031284) on Tuesday July 08, 2008 @10:42AM (#24100061) Homepage
    So it turns something cheap and unreliable into something most people can't afford that doesn't perform well in crashes?
  • The guy creates a photon counter out of a little diode and you're still making fun of your moderators?

    Bad form.

  • Ferrari? Not quite (Score:5, Insightful)

    by bughunter (10093) <bughunter AT earthlink DOT net> on Tuesday July 08, 2008 @10:48AM (#24100159) Journal
    FTFA:

    If you haven't quite seen the significance of this, imagine overclocking your calculator and matching the performance of a workstation. Or polishing up the 3 inch reflector in your attic and outclassing Hubble with your images.

    I'd say it's more like finding out your "workstation" is an overpriced, overcomplicated Rube Goldberg device that in reality has the same performance as a Razor scooter.

    Up to this point, photon counters were elaborate devices with scintillation media, anticoincidence detctors, veto logic, and complex timing and biasing requirements.

    Now you can just apply 9.8V and an instrumentation amp and a couple analog filter/comparator chains, and off you go counting.

    • by mikael (484)

      And the commodity price of photo diodes shoots up to match that of photon counters.

      • Re: (Score:3, Funny)

        by Bluesman (104513)

        That's no joke. I was going to buy a photon counter for my wife and I to share, but now, I'm going to buy about ten of these for each of us, just in case of a hurricane or something.

        Hopefully economies of scale will kick in once these hit the Best Buy shelves.

    • by evanbd (210358) on Tuesday July 08, 2008 @12:38PM (#24101929)

      Up to this point, photon counters were elaborate devices with scintillation media, anticoincidence detctors, veto logic, and complex timing and biasing requirements.

      Now you can just apply 9.8V and an instrumentation amp and a couple analog filter/comparator chains, and off you go counting.

      Single photon avalanche diodes produce rising edge times well under 1ns. You need to measure the *shape* of that rising edge to use this technique. That is a complex circuit, no matter how you look at it.

      The new circuits will be vastly simpler. But they will require a fair bit more than instrument amp and a ballast resistor and a comparator.

    • by Virmal (1281900)
      No wonder everyone was gawking at my "photon enhanced" Fiat 500 the other day...
  • by Dr.Pete (1021137) on Tuesday July 08, 2008 @10:51AM (#24100197) Journal
    This kind of gorgeous tweaking gives me warm feelings inside. Shields has taken a common device used in the field and, through a deep understanding of the physics of its operations, has increased it's functionality without much additional complexity. From the paper he says he cools the device thermo-electrically to -30 deg. C. Thermo-electric cooling is far nicer than cryogenic cooling (typ. using liquid gasses for heat exchange) used in other devices for photon number counting. Further, the method only introduces electronic capacitance subtraction of the photodiode response which is relatively simple compared to other methods (e.g. http://www.stanford.edu/group/fejer/fejerpubs/2005/Langrock_OL_2005.pdf [stanford.edu] which uses the nonlinear response of a crystal and a massive amount of supporting optics and electronics). This subtraction gives orders of magnitude greater sensitivity and allows the time response of the initial avalanche to be extracted from which photon numbers can be counted. One of those wonderful, "why didn't I think of that", insights. Very nice.
    • Re: (Score:2, Interesting)

      by spike_gran (219938)

      This is pretty neat.

      I used avalanche photo diodes as cheap photon counters back in university a decade ago. In our case, however, we would quench the diode after each count and allow it to reset. This works fine except that it severly limits the rate at which photons can be counted and doesn't distinguish when multiple photons arrive simultaneously.

      So this is a cool extension to that technology.

  • They say that in the first instants after the avalanche forms... All you have to do is measure it at this early stage... like turning a Fiat 500 into a Ferrari.

    Summary sounds like a v1agra spam ad...

  • Simple Mod Turns Diodes Into Photon Counters

    and yet he still couldn't get his Lambretta to start...
  • Right. So how long before these simpler devices get into digital cameras? It would certainly help to increase the signal to noise ratio.

    I wonder how these things would perform in direct sunlight. You're getting a few more than a couple of photons at once in that case....

  • The Y axis is probability, the x axis is millivolts. The output of an avalanche photodiode is current. So what is being measured here and how does it relate to the article?

    If this is the probability of finding particular signal amplitudes (time unspecified), how does it relate to single photon counting?

    It seems all back to front. Naively one might expect that multiple photons would result in a larger initial avalanche current, but the graph doesn't seem to relate to this at all. It shows a high probability

    • Re: (Score:3, Informative)

      by argent (18001)

      I think you understand the graph, the presentation is just non-intuitive.

      What the graph is saying is that if 0 photons come in, the maximum probability is around 4.5 millivolts. If 1 photon comes in, the maximum probability is about 7.5 mV, and the high end tail of the curve for the 1 photon case is negligible over 11 mV. If 2 photons come in the peak is at 12 mV, and the low end tail of the curve for 2 photons is negligible at 7.5 mV.

      In which case what is being shown is that the resting output is around 5m

  • Isn't that what those little cells (rods) in our eyes do?

    I remember they can detect one single photon, (although we wouldn't perceive it) going all the way up to super bright light.

    Although they are more of an analog detector, (stronger signal just looks brighter) rather than digital, it's pretty much the same device.

    • by Ihlosi (895663)

      I remember they can detect one single photon, (although we wouldn't perceive it) going all the way up to super bright light.

      The big news is that with the discovery, you can tell the difference between a single photon and two or three of them.

      Detecting that there were any photons isn't the big challenge - determining the exact number of them is.

  • Isn't that what the Starship Enterprise used to destroy enemy ships?

  • Can some less physics challenged person enlighten me as to how we can actually manipulate a single photon. That to me seems to be such a small amount of energy as to be undetectable. I mean if I remember correctly, if you bounce one electron down from one high level to a lower level on one atom, then, it would give off one photon... and getting one atom to do that seems rather a tall challenge..

    • I'm with you... I'm a CS major, not a physics geek, so I'm still in awe that scientists can not only emit and detect individual photos, but do so reliably in several quantum states.

      I thought our use of light and particle streams was still in the "drink from a firehose of similar particles" stage.

    • by moosesocks (264553) on Tuesday July 08, 2008 @03:27PM (#24104551) Homepage

      Believe it or not, "a single photon" isn't as small of an amount of light as you'd think.

      In a study [ucr.edu], 60% of participants were able to correctly identify a pulse of 90 "green" photons. Because only approximately 10% of the light that enters your eye ends up on your retina, that's just 9 photons required to trigger a neural response.

      Because your retinas have approx. 350 rods in them, which sense light in a dark environment (and only in black & white), those 9 photons are spread across those 350, which can be interpreted to mean that parts of your eye are indeed responding to single photons.

      Considering just how small of an amount of light/energy is contained within a single photon, this result is absolutely astonishing.

      For more information regarding single photons, read up on the photoelectric effect [wikipedia.org]. It's quite simple in concept, and its discovery by Einstein in 1905 conclusively confirmed the notion that light exists as a particle.

      This paved the way to Quantum Physics, and won Einstein the Nobel prize in 1921.

      • by ozbird (127571)
        Believe it or not, "a single photon" isn't as small of an amount of light as you'd think.

        Oh? Is there a smaller amount of light than a single photon?
        • by dissy (172727)

          Believe it or not, "a single photon" isn't as small of an amount of light as you'd think.

          Oh? Is there a smaller amount of light than a single photon?

          Half of a photon!

          As it turns out, the government now taxes light, so 50% of it goes right to the IRS, and you only get to keep a smaller amount.

        • by Ihlosi (895663)
          Oh? Is there a smaller amount of light than a single photon?

          A single photon is infinitely more light than no photon.

      • Rods and Cones The retina uses special cells called rods and cones to process light. Just how many rods and cones does your retina have? How about 120 million rods and 7 million cones â" in each eye! http://kidshealth.org/kid/htbw/eyes.html [kidshealth.org]
      • Re: (Score:2, Interesting)

        by Anonymous Coward

        You know how you get that grainy effect in your vision under very low light conditions with well adjusted night vision? That's quantum noise due to individual photons hitting your eye [dspguide.com].

        • Re: (Score:3, Interesting)

          by Raenex (947668)

          You know how you get that grainy effect in your vision under very low light conditions with well adjusted night vision? That's quantum noise due to individual photons hitting your eye [dspguide.com].

          Very interesting. From the link:

          "Vision appears very noisy in near darkness, that is, the image appears to be filled with a continually changing grainy pattern. This results from the image signal being very weak, and is not a limitation of the eye. There is so little light entering the eye, the random detection of individual photons can be seen. This is called statistical noise, and is encountered in all low-light imaging, such as military night vision systems."

      • Wow, your UCR link is old :-)

        First, a couple paragraphs about your parent article:

        The ease with which single photons are emitted varies with wavelength. High energy photons (gamma rays) are fairly straightforward to produce using a radioisotope, spallation or annihilation emitter, a screen opaque to longer-wavelength photons, and a gate (either polarizing or deflecting). Single emissions of lower energy photons (~ 1300-1600 nanometre useful-in-fibre wavelengths, for example, and some visible wavelengths

    • by oojah (113006)

      Generating small numbers of photons is quite commonplace. Fluorescence imaging is often done as single photon or two photon imaging - there's lots if you search for that in google (and pictures too!). It's exactly the kind of thing you describe - excite the atom with your incoming light pulse and then get a different wavelength out when the electons decays to its previous state. You'll likely be looking at using femto-second lasers (ie. pulses a fs long) to do it. I confess this is an area that I'm not part

      • by bh_doc (930270)

        What gets me still is the wave-particle duality of light. Working with photodiodes, I treat light exclusively as photons and everything else just makes my head hurt :)

        Where I work we do experiments where we manipulate single photons with their wave properties. You know- polarization, path splitting (quantum superposition of single-photon paths) and interferometry, phase, interference, etc, all down to the single photon level.

        Ow. *Goes to lie down, again*

      • Two-photon imaging isn't measuring two photons. It's using a laser at half the excitation energy of the dye, so that two photons have to hit the dye almmost simultaneously to excite it. Exciting a dye with near ir instead of blue light has lower efficiency, but it improves resolution dramatically.

        For example, if you are imaging a sample labeled with a fluorescein-type dye, normal imaging would excite with a blue (488) laser line, and emission would be in the green range. If you are imaging a thick spe
        • by oojah (113006)

          Thanks for the clarification. That makes much more sense than the muddled idea I had of it.

  • by Fantastic Lad (198284) on Tuesday July 08, 2008 @03:11PM (#24104305)

    What a strange analogy.

    [. . .] the original Fiat 500, launched in 1957, still holds a place in the heart of most Italians similar to the original Mini for the Brits, the Citroën Deux Chevaux for the French, or the VW Beetle for the Germans (and many North Americans, too).

    Cinquecento was really the car that put Italy on wheels. In short, it's a cultural icon.

    [. . .]
    The original's 500 name came from the displacement of the air-cooled, rear-mounted two-cylinder engine. Slow? Oh my dear God yes.

    But cute as a bunny and tough as nails. You still see these collector's items in every Italian city. I'd have one in a heartbeat.

      --Some car magazine. [wheels.ca]

    I wonder if a classic car buff would really want to convert their beloved collector's piece into a modern consumerist status symbol for power and wealth?

    In any case, while one is fast, and the other is slow, the Fiat was originally made small so that it could navigate the teeny one-lane streets featured in many Italian cities, whereas the Ferrari needs some serious hubcap room. It could be argued that the analogy should be reversed since the smaller car is more agile and able to deal with small details whereas the other cannot and is in fact primarily focused on flying through as many kilometers as quickly as possible and isn't terribly concerned with counting them off in smaller quantities.

    Of course, this kind of observation is the reason why I would be irritating at parties. Carry on. I'm listening.

    -FL

    • Would be irritating at parties? I suppose that we need to apply the scientific principle. We have the hypothesis (irritating at parties), lets test it and revise the hypothesis! (you crash a frat bash and see what happens)
    • by Goldsmith (561202)

      Yeah, it's a bad analogy.

      The opposite would be better:

      Where as previously, you needed a Ferrari to do single photon measurements, now a Fiat will do just as well.

  • You could then measure the distance to the moon with a 10 Watts laser, a telescope, a very precise clock and the newly build photon counter. On the moon are placed some special reflection spots just for exactly this purpose.

    • by 4D6963 (933028)

      You could then measure the distance to the moon with a 10 Watts laser, a telescope, a very precise clock and the newly build photon counter. On the moon are placed some special reflection spots just for exactly this purpose.

      Utter bullcrap. The signal you get back is quadrillion times (I'm not making that up) weaker than the signal you send, to the point that scientists who do that use very powerful lasers (forgot how much, but it's a lot, look it up) and do it for about 10 minutes (again, if I recall correctly) so that they can distinguish their signal from noise. Besides I believe there's already CCDs out there that are noiseless and can count single photons (not in commercial cameras of course).

  • Thank you! (Score:3, Funny)

    by 4D6963 (933028) on Tuesday July 08, 2008 @06:44PM (#24107583)

    That's like turning a Fiat 500 into a Ferrari

    Ha, thank you, dear Slashdot editor, for inserting a necessary car analogy here! As you rightly guessed, it came right at the point in the summary where my feeble mind wondered "huh?!? photon avalanche wtf???".

    Thanks to your enlightening analogy my next thoughts were "ha, a Ferrari, of course! So this has nothing to do with the dangers of skiing on beams of light!", at which point I decided to stop reading the summary as these few concise words seemed to synthesise perfectly the very essence of this discovery, and that my mind proceeded to wander about how awesome it would be to be able to ski on waves of light!

    Thank you a thousand times, oh most esteemed and wise Slashdot editor!

Truly simple systems... require infinite testing. -- Norman Augustine

Working...