MICHAEL GOODMAN Quantum Seeing in the Dark

Quantum optics demonstrates the existence of interaction-free measurements: the detection of objects without light—or anything else—ever hitting them

by Paul Kwiat, Harald Weinfurter and Anton Zeilinger

n Greek mythology, the hero Per- have allowed some bit of light striking field of quantum optics have learned seus is faced with the unenviable Medusa to reflect into his eye; having that not only is this claim far from obvi- I task of fighting the dreaded Medu- thus “seen” the monster, he would have ous, it is, in fact, incorrect. For we now sa. The snake-haired beast is so hideous been finished. know how to determine the presence of that a mere glimpse of her immediately In the world of physics, this predica- an object with essentially no photons turns any unlucky observer to stone. In ment might be summed up by a seem- having touched it. one version of the story, Perseus avoids ingly innocuous, almost obvious claim Such interaction-free measurement this fate by cleverly using his shield to made in 1962 by Nobelist Dennis Ga- seems to be a contradiction—if there is reflect Medusa’s image back to the crea- bor, who invented holography. Gabor no interaction, how can there be a mea- ture herself, turning her to stone. But asserted, in essence, that no observation surement? That is a reasonable conun- what if Perseus did not have well-pol- can be made with less than one pho- drum in classical mechanics, the field of ished armor? He presumably would ton—the basic particle, or quantum, of physics describing the motions of foot- have been doomed. If he closed his light—striking the observed object. balls, planets and other objects that are eyes, he would have been unable to find In the past several years, however, not too small. But quantum mechanics— his target. And the smallest peek would physicists in the increasingly bizarre the science of electrons, photons and

72 Scientific American November 1996 Quantum Seeing in the Dark Copyright 1996 Scientific American, Inc. LASER BEAM following a spiraling path the light so that there is very little chance that leads to a photon detector can illus- of it hitting the pebble. For the player trate the so-called quantum Zeno effect, an to see the pebble, however, at least one element of interaction-free measurements. photon must have hit it, by definition, implying that he has lost. other particles in the atomic realm—says Elitzur, Vaidman and the Bomb otherwise. Interaction-free measure- ments can indeed be achieved by quan- o make the game more dramatic, tum mechanics and clever experimental TAvshalom C. Elitzur and Lev Vaid- designs. If Perseus had been armed with man, two physicists at Tel Aviv Univer- a knowledge of quantum physics, he sity, considered the pebble to be a “su- could have devised a way to “see” Me- perbomb” that would explode if just a dusa without any light actually striking single photon hit it. The problem then the Gorgon and entering his eye. He became: determine if a pebble bomb could have looked without looking. sits under a shell, but don’t set it off. Such quantum prestidigitation offers Elitzur and Vaidman were the first re- many ideas for building detection de- searchers to offer any solution to the INTERFERENCE occurs when a laser is vices that could have use in the real problem. Their answer works, at best, shone through two slits, which generate concentric light waves that interfere with world. Perhaps even more interesting half the time. Nevertheless, it was es- each other (top).The waves can add con- are the mind-boggling philosophical sential for demonstrating any hope at structively or destructively (middle), yield- implications. Those applications and all of winning the game. ing the characteristic interference pattern implications are best understood at the Their method exploits the fundamen- of bright and dark bands (bottom). level of thought experiments: stream- tal nature of light. We have already men- lined analyses that contain all the essen- tioned that light consists of photons, tial features of real experiments but calling to mind a particlelike quality. without the practical complications. But light can display distinctly wavelike LIGHT So, as a thought experiment, consider characteristics—notably a phenomenon SOURCE a variation of a shell game, which em- called interference. Interference is the ploys two shells and a pebble hidden way two waves add up with each other. under one of them. The pebble, howev- For example, in the well-known double- er, is special: it will turn to dust if ex- slit experiment, light is directed through posed to any light. The player attempts two slits, one above the other, to a far- SCREEN to determine where the hidden pebble is away screen. The screen then displays but without exposing it to light or dis- bright and dark fringes [see illustration turbing it in any way. If the pebble turns at right]. The bright fringes correspond CONSTRUCTIVE INTERFERENCE to dust, the player loses the game. to places where the crests and troughs Initially, this task may seem impossi- of the light waves from one slit add con- ble, but we quickly see that as long as structively to the crests and troughs of + = the player is willing to be successful half waves from the other slit. The dark the time, then an easy strategy is to lift bands correspond to destructive inter- the shell he hopes does not contain the ference, where the crests from one slit DESTRUCTIVE INTERFERENCE pebble. If he is right, then he knows the cancel the troughs from the other. An- pebble lies under the other shell, even other way of expressing this concept is though he has not seen it. Winning with to say that the bright fringes correspond + = this strategy, of course, amounts to to areas on the screen that have a high nothing more than a lucky guess. Next, we take our modification one step further, seemingly simplifying the game but in actuality making it impos- INTERFERENCE sible for a player limited to the realm of PATTERN classical physics to win. We have only one shell, and a random chance that a pebble may or may not be under it. The player’s goal is to say if a pebble is pres- ent, again without exposing it to light. Assume there is a pebble under the shell. If the player does not look under the shell, then he gains no information. If he looks, then he knows the pebble SLITS was there, except that he has necessari- ly exposed it to light and so finds only a LIGHT SOURCE

pile of dust. The player may try to dim JARED SCHNEIDMAN DESIGN

Quantum Seeing in the Dark Scientific American November 1996 73 Copyright 1996 Scientific American, Inc. probability of photon hits, and the dark only one way to reach it. Therefore, the fringes to a low probability of hits. photon makes another random choice According to the rules of quantum at the second beam splitter. It may be mechanics, interference occurs whenev- reflected and hit detector D-light; this er there is more than one possible way outcome gives no information, because for a given outcome to happen, and the it would have happened anyway if the

JARED SCHNEIDMAN DESIGN ways are not distinguishable by any pebble had not been there. But the pho- means (this is a more general definition ton may also go to detector D-dark. If PHYSICIST’S SHELL GAME is a thought ex- of interference than is often given in text- that occurs, we know with certainty periment that illustrates the potential of books). In the double-slit experiment, that there was an object in one path of interaction-free measurements. A special light can reach the screen in two possible the interferometer, for if there were not, pebble may be under a shell; if any light ways (from the upper or the lower slit), detector D-dark could not have fired. touches the pebble, it turns to dust. How and no effort is made to determine which And because we sent only a single pho- can one determine which shell hides photons pass through which slit. If we ton, and it showed up at D-dark, it could the pebble? somehow could determine which slit a not have touched the pebble. Somehow photon passed through, there would be we have managed to make an interac- no interference, and the photon could tion-free measurement—we have deter- end up anywhere on the screen. As a re- mined the presence of the pebble with- sult, no fringe pattern would emerge. out interacting with it. Simply put, without two indistinguish- Although the scheme works only some able paths, interference cannot occur. of the time, we emphasize here that when As the initial setup for their hypothet- the scheme works, it works completely. ical measuring system, Elitzur and Vaid- The underlying quantum-mechanical man start with an interferometer—a de- magic in this feat is that everything, in- vice consisting of two mirrors and two cluding light, has a dual nature—both beam splitters. Light entering the inter- particle and wave. When the interfer- ferometer hits a beam splitter, which ometer is empty, the light behaves as a sends the light along two optical paths: wave. It can reach the detectors along ELITZUR-VAIDMAN EXPERIMENT gives a an upper and a lower one. The paths both paths simultaneously, which leads photon a choice of two paths to follow. recombine at the second beam splitter, to interference. When the pebble is in The optical elements are arranged (top) so which sends the light to one of two pho- place, the light behaves as an indivisible that photons always go to detector D-light ton detectors [see illustration at left]. particle and follows only one of the (corresponding to constructive interfer- Thus, the interferometer gives each pho- paths. The mere presence of the pebble ence) but never to D-dark (corresponding ton two possible paths between the light removes the possibility of interference, to destructive interference). The presence source and a detector. even though the photon need not have of a pebble in one path, however, occa- If the lengths of both paths through interacted with it. sionally sends a photon to D-dark (bot- the interferometer are adjusted to be ex- To demonstrate Elitzur and Vaid- tom), indicating that an interaction-free actly equal, the setup effectively becomes man’s idea, we and Thomas Herzog, measurement has occurred. the double-slit experiment. The main now at the University of Geneva, per- difference is that the photon detectors formed a real version of their thought take the place of the screen that shows experiment two years ago and thus dem- D-DARK bright and dark fringes. One detector is onstrated that interaction-free devices D-LIGHT positioned so that it will detect only the can be built. The source of single pho- MIRROR equivalent of the bright fringes of an in- tons was a special nonlinear optical crys- terference pattern (call that detector D- tal. When ultraviolet photons from a BEAM light). The other one records the dark laser were directed through the crystal, SPLITTERS fringes—in other words, no photon ever sometimes they were “down-converted” reaches it (call that detector D-dark). into two daughter photons of lower en- ergy that traveled off at about 30 de- MIRROR Pebble in the Path grees from each other. By detecting one of these photons, we were absolutely hat happens if a pebble is placed certain of the existence of its sister, which Winto one of the paths, say, the up- we then directed into our experiment. D-DARK per one? Assuming that the first beam That photon went into an interfer- splitter acts randomly, then with 50 ometer (for simplicity, we used a slight- PEBBLE D-LIGHT percent likelihood, the photon takes the ly different type of interferometer than upper path, hits the pebble (or explodes the one Elitzur and Vaidman proposed). the superbomb) and never gets to the The mirrors and beam splitter were second beam splitter. aligned so that nearly all the photons left If the photon takes the lower path, it by the same way they came in (the ana- does not hit the pebble. Moreover, in- logue of going to detector D-light in the terference no longer occurs at the sec- Elitzur-Vaidman example or, in the dou-

JARED SCHNEIDMAN DESIGN ond beam splitter, for the photon has ble-slit experiment, of going to a bright

74 Scientific American November 1996 Quantum Seeing in the Dark Copyright 1996 Scientific American, Inc. fringe). In the absence of the pebble, the DETECTOR chance of a photon going to detector D- dark was very small because of destruc- D-DARK tive interference (the analogue of the dark fringes in the double-slit experi- DOWN-CONVERSION ment) [see illustration at right]. CRYSTAL BEAM SPLITTER But introducing a pebble into one of the pathways changed the odds. The MIRROR pebble was a small mirror that directed the light path to another detector (D- MIRROR pebble). We then found that about half (PEBBLE) of the time, D-pebble registered the pho- ton, whereas about one fourth of the time D-dark did (the rest of the time the photon left the interferometer the same DETECTOR way it came in, giving no information). The firing of D-dark was the interac- MIRROR tion-free detection of the pebble. JARED SCHNEIDMAN DESIGN In a simple extension of the scheme, we reduced the reflectivity of the beam ments can have on quantum systems. D-DARK splitter, which lessened the chance that The phenomenon is called the quantum the photons would be reflected onto the Zeno effect, because it resembles the fa- path containing the mirror to D-pebble. mous paradox raised by the Greek phi- BEAM SPLITTER What we found, in agreement with the- losopher Zeno, who denied the possi- oretical prediction, was that the proba- bility of motion to an arrow in flight bilities of the photons going to D-peb- because it appears “frozen” at each in- ble and going to D-dark became more stant of its flight. It is also known as the and more equal. That is, by using a bare- watched-pot effect, a reference to the MIRROR ly reflective beam splitter, up to half the aphorism about boiling water. We all (PEBBLE) measurements in the Elitzur-Vaidman know that the mere act of watching the scheme can be made interaction-free (in- pot should not (and does not) have any DETECTOR stances in which the photons leave the effect on the time it takes to boil the MIRROR interferometer the same way they came water. In quantum mechanics, however, in are not counted as measurements). such an effect actually exists—the mea- surement affects the outcome (the prin- DEMONSTRATION of the Elitzur-Vaidman The Quantum Zeno Effect ciple is called the projection postulate). scheme uses light from a down-conver- Kasevich essentially reinvented the sion crystal, which enters a beam splitter, he question immediately arose: Is 50 simplest example of this effect, which bounces off two mirrors and interferes Tpercent the best we can do? Consid- was first devised in 1980 by Asher Pe- with itself back at the beam splitter (top). erable, often heated, argument ensued res of the Technion-Israel Institute of No light reaches D-dark (corresponding among us, for no design change that Technology. The example exploits yet to destructive interference; constructive would improve the odds was evident. another characteristic of light: polariza- interference is in the direction from which In January 1994, however, Mark A. tion. Polarization is the direction in the photon first came). If a mirror “pebble” Kasevich of Stanford University came which light waves oscillate—up and is inserted into a light path, no interfer- to visit us at Innsbruck for a month, down for vertically polarized light, side ence occurs at the beam splitter; D-dark and during this stay he put us on to a to side for horizontally polarized light. sometimes receives photons (bottom). solution that, if realized, makes it possi- These oscillations are at right angles to ble to detect objects in an interaction- the light’s direction of propagation. Light free way almost every time. It was not from the sun and other typical sources POLARIZATION refers to the vibrations of the first instance, and hopefully not the generally vibrates in all directions, but light waves as they move through space. last, in which quantum optimism tri- umphed over quantum pessimism. The new technique is more or less an application of another strange quantum phenomenon, first discussed in detail in 1977 by Baidyanath Misra, now at the University of Brussels, and E. C. George Sudarshan of the University of Texas at Austin. Basically, a quantum system can be trapped in its initial state, even though it would evolve to some other state if ORDINARY VERTICALLY HORIZONTALLY left on its own. The possibility arises be- LIGHT POLARIZED LIGHT POLARIZED LIGHT

cause of the unusual effect that measure- JARED SCHNEIDMAN DESIGN

Quantum Seeing in the Dark Scientific American November 1996 75 Copyright 1996 Scientific American, Inc. does the trick. Here’s why: After the first rotator, the light is not too much turned from the horizontal. This means that the chance that the photon is ab- POLARIZATION ROTATORS HORIZONTAL POLARIZERS sorbed in the first horizontal polarizer is quite small, only 6.7 percent. (Math- ematically, it is given by the square of the sine of the turning angle.) If the photon is not absorbed in the JARED SCHNEIDMAN DESIGN first polarizer, it is again in a state of horizontal polarization—it must be, be- QUANTUM ZENO EFFECT can be demon- here we are concerned mostly with cause that is the only possible state for strated with devices that rotate polariza- vertical and horizontal polarizations. light that has passed a horizontal polar- tion 15 degrees. After passing through six Consider a photon directed through izer. At the second rotator, the polariza- such rotators, the photon changes from a a series of, say, six devices that each tion is once again turned 15 degrees horizontal polarization to a vertical one slightly rotates the polarization of from the horizontal, and at the second and so is absorbed by the polarizer (top light so that a horizontally polarized polarizer, it has the same small chance row). Interspersing a polarizer after each photon ends up vertically polarized of being absorbed; otherwise, it is again rotator, however, keeps the polarization [see illustration above]. These rota- transmitted in a state of horizontal po- from turning (bottom row). tors might be glass cells containing larization. The process repeats until the sugar water, for example. At the end photon comes to the final polarizer. of the journey through the rotators, An incident photon has a two-thirds the photon comes to a polarizer, a chance of being transmitted through all device that transmits photons with six inserted polarizers and making it to EXPERIMENTAL REALIZATION of the quan- one kind of polarization but absorbs the detector; the probability is given by tum Zeno effect was accomplished by photons with perpendicular polar- the relation (cos2(15 degrees))6 . Yet as making the photon follow a spiral-stair- ization. In this thought experiment, we increase the number of stages, de- case path, so that it traversed the polariza- the polarizer transmits only horizon- creasing the polarization-rotation angle tion rotator six times. Inserting a polarizer tally polarized light to a detector. at each stage accordingly (that is, 90 de- next to the rotator suppressed the rota- We will start with a photon hori- grees divided by the number of stages), tion of the photon’s polarization. zontally polarized, and each rotator the probability of transmitting the pho- will turn the polarization by 15 de- ton increases. For 20 stages, the proba- grees. It is clear, then, that the pho- bility that the photon reaches the detec- ton will never get to the detector, for tor is nearly 90 percent. If we could make after passing through all the cells, its a system with 2,500 stages, the proba- polarization will have turned 90 de- bility of the photon being absorbed by grees (15 degrees for each of the six one of the polarizers would be just one rotators) so that it becomes vertical. in 1,000. And if it were possible to have uck The polarizer absorbs the photon. an infinite number of stages, the photon nnsbr This stepwise rotation of the polar- would always get through. Thus, we y of I

ersit ization is the quantum evolution that would have completely inhibited the niv

U we wish to inhibit. evolution of the rotation. Interspersing a horizontal polariz- To realize the quantum Zeno effect, er between each polarization rotator we used the same nonlinear crystal as

MICHAEL RECK before to prepare a single photon. In- stead of using six rotators and six po- larizers, we used just one of each; to HORIZONTALLY POLARIZED achieve the same effect, we forced the POLARIZATION photon through them six times, employ- PHOTONS ROTATOR INSERTABLE POLARIZING ing three mirrors as a kind of spiral MIRROR BEAM SPLITTER staircase [see illustration at left]. In the absence of the polarizer, the photon ex- iting the staircase is always found to be vertically polarized. When the polarizer MIRROR is present, we found that the photon was horizontally polarized (unless the polarizer blocked it). These cases oc- POLARIZER curred roughly two thirds of the time for MIRROR DETECTOR our six-cycle experiment, as expected from our thought-experiment analysis. Next we set out to make an interac- tion-free measurement—that is, to detect

JARED SCHNEIDMAN DESIGN an opaque object without any photons

76 Scientific American November 1996 Quantum Seeing in the Dark Copyright 1996 Scientific American, Inc. hitting it—in a highly efficient manner. we can make the probability that the We devised a system that was somewhat photon is absorbed by the object as small of a hybrid between the Zeno example as we like. Preliminary results from new and the original Elitzur-Vaidman meth- experiments at Los Alamos National od. A horizontally polarized photon is Laboratory have demonstrated that up let into the system and makes a few cy- to 70 percent of measurements could MIRROR cles (say, six again) before leaving. (For be interaction-free. We soon hope to in- this purpose, one needs a mirror that can crease that figure to 85 percent. be “switched” on and off very quickly; PEBBLE OMETER fortunately, such mirrors, which are ac- Applying Quantum Magic tually switchable interference devices, POLARIZING BEAM INTERFER have already been developed for pulsed hat good is all this quantum con- SPLITTER lasers.) At one end of the system is a Wjuring? We feel that the situation MIRROR polarization rotator, which turns the resembles that of the early years of the photon’s polarization by 15 degrees in laser, when scientists knew it to be an each cycle. The other end contains a po- ideal solution to many unknown prob- POLARIZATION larization interferometer. It consists of a lems. The new method of interaction- ROTATOR polarizing beam splitter and two equal- free measurement could be used, for in- SWITCHABLE length interferometer paths with mirrors stance, as a rather unusual means of MIRROR at the ends [see illustration at right]. photography, in which an object is im- JARED SCHNEIDMAN DESIGN At the polarizing beam splitter, all aged without being exposed to light. horizontally polarized light is transmit- The “photography” process would EFFICIENT MEASUREMENTS that are inter- ted, and all vertically polarized light is work in the following way: Instead of action-free combine the setups of the reflected; in essence, the transmission sending in one photon, we would send quantum Zeno effect and the Elitzur-Vaid- and reflection choices are analogous to in many photons, one per pixel, and man scheme. The photon enters below the two paths in the double-slit experi- perform interaction-free measurements the switchable mirror and follows the op- ment. In the absence of an object in the with them. In those regions where the tical paths six times before being allowed polarization interferometer, light is split object did not block the light path of to exit through the mirror. Its final polar- at the beam splitter according to its po- the interferometer, the horizontal polar- ization will still be horizontal if there is a pebble in one light path; otherwise, it will larization, reflects off the mirrors in each ization of the photons would undergo have rotated to a vertical polarization. path and is recombined by the beam the expected stepwise rotation to verti- splitter. As a result, the photon is in ex- cal. In those regions where the object actly the same state as before it entered blocked the light path, a few of the pho- the interferometer (that is, with a polar- tons would be absorbed; the rest would ization turned 15 degrees toward the have their polarizations trapped in the vertical). So, after six cycles, the polar- horizontal state. Finally, we would take ization ends up rotated to vertical. a picture of the photons through a po- The situation changes when an opaque larizing filter after they had made the The Projection Postulate object is placed in the vertical polariza- requisite number of cycles. tion path of the interferometer. This sit- If the filter were horizontally aligned, The postulate states that for any measure- uation is analogous to having the six we would obtain an image of the ob- ment made on a quantum system only polarizers inserted in the quantum Zeno ject; if vertically aligned, we would ob- certain answers are possible. Moreover, af- effect experiment. So in the first cycle, tain the negative. In any case, the pic- ter the measurement, the quantum sys- the chance that the photon—the polar- ture is made by photons that have nev- tem is in a state determined by the ob- ization of which has been turned only er touched the object. These techniques tained results. So a photon that has passed through a horizontal polarizer is necessar- 15 degrees from horizontal—enters the can also work with a semitransparent vertical-polarization path (and is then object and may possibly be generalized ily horizontally polarized, even if it were absorbed by the object) is very small to find out an object’s color (although originally polarized at a nearly vertical an- gle (the polarizer eliminates the vertical (6.7 percent, as in the Zeno thought ex- these goals would be more difficult). component of the polarization). The prob- periment). If this absorption does not A variation of such imaging could ability of transmission in this case, though, happen, the photon must have entered someday conceivably prove valuable in would be low. the horizontal path instead, and its po- medicine—for instance, as a means to larization is reset to be purely horizontal. image living cells. Imagine being able to Just as in the Zeno example, the whole x-ray someone without exposing them process repeats at each cycle, until final- to many penetrating x-rays. Such imag- ly, after six cycles, the bottom mirror is ing would therefore pose less risk to pa- switched off, and the photon leaves the tients than standard x-rays. (Practically system. Measuring the photon’s polar- speaking, such x-ray photography is un- ization, we find it still to be horizontal, likely to be realized, considering the dif- implying that a blocker must reside in ficulty of obtaining optical elements for the interferometer. Otherwise, the pho- this wavelength of light.) ton would have been vertically polarized A candidate for more immediate ap- POLARIZER

when it left. And by using more cycles, plication is the imaging of the clouds of JARED SCHNEIDMAN DESIGN

Quantum Seeing in the Dark Scientific American November 1996 77 Copyright 1996 Scientific American, Inc. of Schrödinger’s cat—a “kitten”—with MIRROR a beryllium ion. They used a combina- MIRROR MEDUSA tion of lasers and electromagnetic fields to make the ion exist simultaneously in two places spaced 83 nanometers apart— a vast distance on the quantum scale. SWITCHABLE If such an ion were interrogated with MIRROR the interaction-free methods, the inter- rogating photon would also be placed POLARIZING in a superposition. It could end up be- BEAM SPLITTER ing horizontally and vertically polarized at the same time. In fact, the kind of ex- INTERACTION-FREE POLARIZATION ROTATOR perimental setup discussed above should PHOTOGRAPHS be able to place a group of, say, 20 pho- tons in the same superposition. Every photon would “know” that it has the POLARIZING same polarization as all the others, but BEAM SPLITTER none would know its own polarization. They would remain in this superposition JARED SCHNEIDMAN DESIGN until a measurement revealed them to be all horizontally polarized or all verti- PHOTOGRAPHY can also be done with in- ultracold atoms recently produced in cally polarized. The sizable bunch of teraction-free techniques. In this way, the various laboratories. The coldest of photons stuck in this peculiar condition object—a “Medusa” that must not be these exhibit Bose-Einstein condensa- would show that quantum effects can viewed directly—will absorb very few tion, a new type of quantum state in be manifested at the macroscopic scale. photons. which many atoms act collectively as Lying beyond the scope of everyday one entity. In such a cloud every atom is experience, the notion of interaction- so cold—that is, moving so slowly—that free measurements seems weird, if not a single photon can knock an atom out downright nonsensical. Perhaps it would of the cloud. Initially, no way existed to seem less strange if one kept in mind get an image of the condensate without that quantum mechanics operates in destroying the cloud. Interaction-free the realm of potentialities. It is because measurement methods might be one way there could have been an interaction that to image such a collection of atoms. we can prevent one from occurring. Besides imaging quantum objects, in- If that does not help, take comfort in teraction-free procedures could also the fact that, over the years, even physi- make certain kinds of them. Namely, cists have had a hard time accepting the the techniques could extend the cre- strangeness of the quantum world. The ation of “Schrödinger’s cat,” a much underlying keys to these quantum feats loved theoretical entity in quantum me- of magic—the complementary, wave- chanics. The quantum feline is prepared and-particle aspect of light and the na- so that it exists in two states at once: it ture of quantum measurements—have is both alive and dead at the same been known since 1930. Only recently time—a superposition of two states. Ear- have physicists started to apply these lier this year workers at the National ideas to uncover new phenomena in Institute of Standards and Technology quantum information processing, in- managed to create a preliminary kind cluding the ability to see in the dark. SA

The Authors Further Reading

PAUL KWIAT, HARALD WEINFURTER and ANTON ZEILINGER QED: The Strange Theory of Light and Matter. Richard freely interacted with one another at the University of Innsbruck. Kwiat, P. Feynman. Princeton University Press, 1985. now a J. R. Oppenheimer Fellow at Los Alamos National Laboratory, Quantum Mechanical Interaction-Free Measurements. earned his Ph.D. from the University of California, Berkeley. He is a seri- Avshalom C. Elitzur and Lev Vaidman in Foundations of ous student of aikido and is trying to become a tolerable flautist. Wein- Physics, Vol. 23, No. 7, pages 987–997; July 1993. furter received his Ph.D. from the Technical University of Vienna and held Interaction-Free Measurement. P. G. Kwiat, H. Weinfur- a postdoctoral position at the Hahn-Meitner Institute in Berlin. He cur- ter, T. Herzog, A. Zeilinger and M. A. Kasevich in Physical rently enjoys all the benefits and comforts of a fellowship from the Austri- Review Letters, Vol. 74, No. 24, pages 4763–4766; June 12, an Academy of Science as well as the relaxing lifestyle in Innsbruck and its 1995. surrounds. A member of the Austrian Academy of Sciences, Zeilinger Discussions on experiments for interaction-free measurements earned his doctorate at the University of Vienna and has held numerous can be found on the World Wide Web at http://info.uibk.ac.at/ appointments worldwide. In his little free time, he plays the double bass c/c7/c704/qo/photon/#Inter and at http://p23.lanl. gov/Quan- and collects antique maps, particularly of the Austro-Hungarian empire. tum/kwiat/ifm-folder/ifmtext.htm

78 Scientific American November 1996 Quantum Seeing in the Dark Copyright 1996 Scientific American, Inc.