Nowadays there are many different ways to communicate. One thing common to all those methods, from carrier pigeon to email, is that the transmission of information requires sending the physical system that represents it from one party to another. If Bob wants to send a message to Alice, then photons, sound waves, pigeons, or some other information-carrying entities need to make the trip. Information is an abstract mathematical concept, but as Rolf Landauer famously pointed out, there is no information without representation.
In the quantum world, however, another option exists: Bob can send a message to Alice, via a kind of carrier pigeon, without having the pigeon actually fly between them. The weird scenario is appropriately called counterfactual communication.
In a recent paper in Proceedings of the National Academy of Sciences, a group headed by Jian-Wei Pan at the University of Science and Technology of China reported an experimental demonstration of counterfactual communication. The researchers sent data with single photons, but the photons did not fly through the communication channel.
The new experiment verifies an earlier theoretical proposal and proves that the predictions of quantum theory hold even in this seemingly impossible task. However, the success of the experiment does not mean that everyone is persuaded that two parties can communicate without exchanging physical messengers (see figure 1). Counterfactual communication has a long history of intense debate, one that is intertwined with enduring disagreements over fundamental interpretations of quantum mechanics.
Explosive origins
Counterfactual communication dates back to a 1993 thought experiment. Avshalom Elitzur and Lev Vaidman wanted to find a way to remotely detect a very dangerous bomb without detonating it. The bomb they envisioned was so sensitive that hitting it with any kind of particle would make it explode. In classical physics, there is no way to detect the bomb without its exploding.
Elitzur and Vaidman’s quantum solution was to send single photons into an interferometer that was aligned so that all the photons would exit the same output port, say A. The appearance of an obstruction, such as a bomb, in one of the arms of the interferometer would break the interference. In that case a photon would have a chance of exiting the other output, B, as shown in figure 2. A click in detector B would herald the presence of the bomb.
Two years later, Paul Kwiat, Anton Zeilinger, and colleagues experimentally verified the proposal. The researchers also suggested a way to increase the efficiency of the scheme using the quantum Zeno effect, by which one can inhibit the evolution of a quantum state over time by measuring it frequently. Instead of having one interferometer, it is better to have a cascade of many interferometers with highly reflective half mirrors that weakly probe the bomb. As the number of interferometers tends toward infinity, the probability of detecting the bomb without exploding it tends toward 1. Clicks in A or B unambiguously correspond to the absence or presence of the bomb inside the apparatus.
Enter the quantum messenger
One of the evolutions of the Elitzur–Vaidman bomb idea was the notion of counterfactual communication. The Zeno-upgraded setup can immediately be used for data transfer. We give the mirrors in one of the interferometer arms (the so-called channel arm) to a distant party, traditionally called Bob. Alice then sends photons into the interferometer, one at a time, and Bob blocks or unblocks his mirrors. By doing so, Bob can transmit a binary message by controlling which detector clicks back at Alice’s side for each photon.
The method is not fully counterfactual, however. Counterfactuality requires that neither of the clicks in Alice’s detectors involves the transfer of a photon between Alice and Bob. In the cases when Bob unblocks his mirror, it’s possible for a photon to propagate through the channel arm. Unlike in the bomb setup, the photon isn’t allowed to pass through the interferometer arm even if the bomb (or in this case, the mirror block controlled by Bob) isn’t there.
A scheme for true counterfactual communication was proposed in 2013 at Texas A&M University by Muhammad Suhail Zubairy and coworkers. They took the complexity of the setup one step further: The cascade of highly reflective interferometers provides the data transfer, but each channel arm is now made of another cascade of highly reflective interferometers. The additional interferometers guarantee, through the Zeno effect, that any part of a photon’s wavefunction that ends up in the channel arms can evolve only toward an auxiliary detector and never toward any of Alice’s detectors. In other words, if there is an occasional photon in Bob’s arm, it will necessarily be detected and discarded so that it does not take part in the data transfer.
In their recent work, Pan and colleagues experimentally verified the Zubairy team’s scheme. The experimentalists managed to send a black-and-white image between two parties; yet, according to their test, a photon was physically detected in the channel arm in only 1.4% of the cases when data were transmitted.
Pan’s team did not have to physically assemble the separate steps. As each step in the scheme is identical to the previous one, they can all be nicely folded into a pair of actively switchable Michelson interferometers—a photon bounces back and forth between the same few mirrors until a switch sends it out (see figure 3). The experiment is still technically challenging. Among many other details, the performance of the scheme depends on the quality of the interference, which has to be kept under strict control via active phase locking and feedback.
Tracking the pigeon
Does the experiment of Pan and colleagues now close the topic? Is it now time to build the counterfactual internet? Probably not. Vaidman and other physicists in his camp have questioned the counterfactuality of the Zubairy scheme in the case of an unblocked channel arm. They ask how one can prove that the photon was never where it was not supposed to be. Their critique harks to a larger issue: What does it mean that “a photon was somewhere”? For classical bodies, like pigeons, we can use a GPS tracker to follow their trajectory as they deliver a letter. Things are not so simple for quantum pigeons.
Physicists on Zubairy’s side argue that the propagation of a photon is described by a wavefunction. If in some section of the path the wavefunction drops to zero, then the photons cannot “flow” through there. And if photons were somehow detected on the other side of the section, they had to have taken a different route. Essentially, Zubairy and others are asserting that a previous path can be assigned to a propagating photon once it is detected. That thinking is more intuitive than Vaidman’s, but quantum mechanics and wavefunctions do not care about our intuition and common sense. As scientists suspected for quite some time and are now quite sure, the quantum world cannot be simultaneously real and local (see Physics Today, January 2016, page 14).
Vaidman, who argues against such defining of quantum trajectories, proposes his own approach to a photon’s past: look for the weak trace it leaves. Finding that trace requires many gentle but inaccurate weak measurements, which barely disturb the system. A weak-trace test of the counterfactual communication scheme shows that before being detected by Alice, the photons left some trace inside the channel arms, but not at those sections that connect the arms to the rest of the setup. That means, according to Vaidman, that the photons just suddenly appeared, propagated a bit, and then vanished without leaving a further trace, like the infamous Cheshire cat of a certain psychedelic wonderland. It also means that the full counterfactuality is not respected.
Not everyone is on board with relying on weak measurements to evaluate counterfactuality. Recovering information still disturbs the system and affects the interference, even if a measurement is weak. In the quantum world, a disturbed system is not the same as an undisturbed one, and the counterfactuality would be broken by the measurement itself. More generally, weak measurements and weak values are themselves the subject of scientific debate. Although the observations they predict can all be derived with the usual quantum mechanics language, the interpretation they provide almost always shows unusual things. In fact, it is not the first time some kind of “Cheshirianity” is observed when the language of weak values is used.
Quantum always wins
The experiment of Pan and his group demonstrates high fidelity of data transmission. From the characterization of their apparatus, the interference visibility, and the single photon source performance, they could concretely conclude that the amount of information going to Alice is much larger than what can be carried by the number of photons detected in Bob’s channel. No doubt there will be ongoing debate about how one measures a photon’s presence at Bob, but really it is an unspeakable question of quantum information. Acquiring this information necessarily disturbs the system.
No matter which interpretation we choose, the experimental demonstration shows a way of transmitting data that relies on a unique quantum feature. Communication with vanishing and reappearing data carriers goes just as much against common sense as communication where the carriers are not even transmitted between the parties.
Sergei Slussarenko, Nora Tischler, and Geoff Pryde are physicists at the Centre for Quantum Computation and Communication Technology and the Centre for Quantum Dynamics at Griffith University in Brisbane, Australia.
