If a tree falls in a forest and no one is there to hear it, does it make a noise? Of course it does, as far as I'm concerned. That is, an observer does not affect reality by merely observing—at least not when it comes to big things. But what about small things, like photons and electrons? Recent work at the Weizmann Institute of Science in Israel indicates that "watching"—at least at scales smaller than one micron—can indeed affect reality.
The work at Weizmann is the latest development in a story that begins in 1801, when the British scientist Thomas Young conducted his two-slit interference experiment. He let filtered sunlight pass through a pinhole and then onto a surface with two more pinholes. The light passed through the two pinholes and to a screen, where the two sources of light generated so-called interference fringes, alternating stripes of brightness and darkness. The bright stripes developed where both waves were at a peak or a trough, and dark stripes indicated where a peak of one wave coincided with a trough of another. Such an example of interference can be seen wherever two waves overlap in space and time, whether the waves exist in light, water or virtually anything else. That experiment might have convinced most scientists that light travels as waves, except for one problem: When Isaac Newton published Optiks in 1704, he described his corpuscular theory of light, which claimed that light travels as particles. Unfortunately for Young, many scientists still accepted Newton's theory.
Eventually, scientists realized that Young was right. Light travels in waves. But in the early 1900s Albert Einstein confused the issue again. The photoelectric effect, where light landing on certain metal surfaces causes electrons to be released, makes sense only if light travels in discrete particles, which Einstein called "light quanta," and which are now called photons. Instead of continuing to argue—particles, waves, particles—scientists concluded that light travels in waves in some circumstances and in particles in others, which led to the concept of wave-particle duality.
Just three more twists set the stage for the work at Weizmann. First, wave-particle duality applies to more than light. For instance, electrons act like particles in some cases and waves in others. Second, Niels Bohr posed the principle of complementarity, which states that something cannot act simultaneously like a particle and a wave. It may act like one or the other, but never like both at the same time. Finally, a contingent of physicists led by Bohr believed that the characteristics of electrons, light and other things at the submicron level depended to some degree on whether anyone tried to observe or measure them. That idea came directly from quantum mechanics, which is essentially a set of principles that explains what submicron-size particles do. Einstein led an opposing group that believed that observation plays no part in determining reality. Things are as they are, Einstein might have said, whether you look at them or not. Consequently, Einstein believed that the theory behind quantum mechanics was incomplete. Bohr and Einstein could only speculate about who was right. Today, physicists can test the two hypotheses.
One important result came in 1991 when Leonard Mandel and his colleagues at the University of Rochester published an optical-interference experiment in Physical Review Letters (67(3):318–321). In essence, they generated two beams of light derived from a laser and caused them to interfere, which is a wave-like phenomenon. But if they used an arrangement that allowed the path of the light to be determined, the interference disappeared, whether or not this determination was actually made. This experiment brought out the particle nature of the light, so by the principle of complementarity it could not simultaneously display the wave-like nature required for interference. "In so far as the experiment confirms the quantum prediction," Mandel says, "the experiment is hardly revolutionary. Perhaps its most useful feature is the demonstration that what matters is not what is known, but what is knowable in principle, because this takes any anthropomorphism out of physics."
But how finely tuned is this principle of complementarity? If you could adjust how closely you looked at a particle, would the level of interference vary accordingly? That's what Mordehai Heiblum and his colleagues at Weizmann wanted to know, and they recently published their findings in Nature (391:871–874). Using state-of-the-art integrated-circuit technology, they built a tiny device—less than 1 micrometer in size—that creates a stream of electrons that pass through a barrier by going down one of two passageways, much like Young's two-slit experiment. The passageways "focus" both electron streams on a device that measures the level of interference between the streams.
Eyal Buks, a graduate student in Heiblum's laboratory and lead author of the article, explains that making such a device work properly requires considerable tuning of the elements or adjusting their voltages. "What makes the tuning problem even more difficult is the tendency of these devices to drift with time." One of the tiny pathways included a particle detector that could be adjusted for sensitivity. To make sure that the detector did not disrupt the flow of electrons along either path, they disconnected one path and looked at the flow of electrons in the other path with the detector on and off. Then they disconnected that path, reconnected the other and looked at the flow of electrons along it with the detector on and off. They found no differences in any of these conditions.
With both paths connected, the electrons interfered—until the investigators turned on the detector. Then the interference disappeared. More interesting still, Buks and his colleagues measured the interference between the two streams of electrons while varying the sensitivity of the detector. When they increased the detector's sensitivity, making it more likely to detect a passing electron, the interference diminished. In other words, more observation led to less interference.
According to Ady Stern, one of several physicists whose theoretical work provided some of the impetus for the recent experiments at Weizmann, "I think the primary significance is having an experiment in which the effective interaction with the environment is so nicely controllable. Here they really have a knob controlling the interaction strength." Such fine control may one day be used in electronics that use an electron's particle nature for some tasks and its wave nature for others. By then, it might seem obvious that an electron's flow through an integrated-circuit forest depends on how many other electrons are "watching."—Mike May