In the anti-laser, incoming light waves are trapped in a cavity where they bounce back and forth until they are eventually absorbed. Their energy is dissipated as heat. (Photo: Yidong Chong)
Scientists at Yale University have built the world’s first anti-laser, in which incoming beams of light interfere with one another in such a way as to perfectly cancel each other out. The discovery could pave the way for a number of novel technologies with applications in everything from optical computing to radiology.
In a paper published in the journal Science they demonstrated that the anti-laser could adsorb 99.4 per cent of incoming light, for a specific wavelength. In theory 99.999% of the incoming light can be absorbed.
In the time-reversed counterpart to laser emission, incident coherent optical fields are perfectly absorbed within a resonator that contains a loss medium instead of a gain medium. The incident fields and frequency must coincide with those of the corresponding laser with gain. We demonstrated this effect for two counterpropagating incident fields in a silicon cavity, showing that absorption can be enhanced by two orders of magnitude, the maximum predicted by theory for our experimental setup. In addition, we showed that absorption can be reduced substantially by varying the relative phase of the incident fields. The device, termed a “coherent perfect absorber,” functions as an absorptive interferometer, with potential practical applications in integrated optics.
Last summer, Yale physicist A. Douglas Stone and his team published a study explaining the theory behind an anti-laser, demonstrating that such a device could be built using silicon, the most common semiconductor material. But it wasn’t until now, after joining forces with the experimental group of his colleague Hui Cao, that the team actually built a functioning anti-laser, which they call a coherent perfect absorber (CPA).
The team, whose results appear in the Feb. 18 issue of the journal Science, focused two laser beams with a specific frequency into a cavity containing a silicon wafer that acted as a “loss medium.” The wafer aligned the light waves in such a way that they became perfectly trapped, bouncing back and forth indefinitely until they were eventually absorbed and transformed into heat.
Stone believes that CPAs could one day be used as optical switches, detectors and other components in the next generation of computers, called optical computers, which will be powered by light in addition to electrons. Another application might be in radiology, where Stone said the principle of the CPA could be employed to target electromagnetic radiation to a small region within normally opaque human tissue, either for therapeutic or imaging purposes.
Theoretically, the CPA should be able to absorb 99.999 percent of the incoming light. Due to experimental limitations, the team’s current CPA absorbs 99.4 percent. “But the CPA we built is just a proof of concept,” Stone said. “I’m confident we will start to approach the theoretical limit as we build more sophisticated CPAs.” Similarly, the team’s first CPA is about one centimeter across at the moment, but Stone said that computer simulations have shown how to build one as small as six microns (about one-twentieth the width of an average human hair).
The team that built the CPA, led by Cao and another Yale physicist, Wenjie Wan, demonstrated the effect for near-infrared radiation, which is slightly “redder” than the eye can see and which is the frequency of light that the device naturally absorbs when ordinary silicon is used. But the team expects that, with some tinkering of the cavity and loss medium in future versions, the CPA will be able to absorb visible light as well as the specific infrared frequencies used in fiber optic communications.
It was while explaining the complex physics behind lasers to a visiting professor that Stone first came up with the idea of an anti-laser. When Stone suggested his colleague think about a laser working in reverse in order to help him understand how a conventional laser works, Stone began contemplating whether it was possible to actually build a laser that would work backwards, absorbing light at specific frequencies rather than emitting it.
“It went from being a useful thought experiment to having me wondering whether you could really do that,” Stone said. “After some research, we found that several physicists had hinted at the concept in books and scientific papers, but no one had ever developed the idea.”
A laser beam from a tunable (800 to 1000 nm) continuous-wave Ti:sapphire source enters a beam splitter (designated 1). The two split beams are directed normally onto opposite sides of a silicon wafer of thickness ~110 μm, using a Mach-Zehnder geometry. A phase delay in one of the beam paths controls the relative phase of the two beams. An additional attenuator ensures that the input beams have equal intensities, compensating for imbalances in the beam splitters and other imperfections. The output beams are rerouted, via beam splitters (designated 2, 3, and 4), into a spectrometer. The inset is a schematic of the CPA mechanism. The incident beams from left and right multiply scatter within the wafer with just the right amplitude and phase so that the total transmitted and reflected beams destructively interfere on both sides, leading to perfect absorption.
Phase modulation of beam absorption. (A) Theoretical plot of normalized total output intensities as a function of wavelength λ for parity-odd (blue) and parity-even (red) scattering eigenmodes. The dashed black line is the result for incoherent input beams. (B to D) Theoretical output intensities at three representative values of λ as the relative phase of the input beams is varied, showing intensities emitted to the right (magenta) and left (green) sides of the slab, and the total intensity (black). Values of λ corresponding to (B) to (D) are marked by vertical lines in (A); (B) is the CPA resonance. (E to G) Experimental results at values of λ approximately corresponding to (B) to (D). Solid lines are fits to the data, not theory curves; results are normalized to max(Iout) of the fit.