Findings of “greatly enhanced” optics turn out to be, well, greatly enhanced

nature photonicsThe authors of a paper in Nature Photonics have been forced to walk back their article after learning from another group of researchers that their conclusions likely were an, ahem, optical illusion.

The paper, “Greatly enhanced continuous-wave terahertz emission by nano-electrodes in a photoconductive photomixer,” appeared in January 2012 and came from a team led by that included Aaron Danner, an optics expert at the National University of Singapore. As the abstract of the paper explains (to physicists, anyway):

An efficient, room-temperature-operation continuous-wave terahertz source will greatly benefit compact terahertz system development for high-resolution terahertz spectroscopy and imaging applications. Here, we report highly efficient continuous-wave terahertz emission using nanogap electrodes in a photoconductive antenna-based photomixer. The tip-to-tip nanogap electrode structure provides strong terahertz field enhancement and acts as a nano-antenna to radiate the terahertz wave generated in the active region of the photomixer. In addition, it provides good impedance-matching to the terahertz planar antenna and exhibits a lower RC time constant, allowing more efficient radiation, especially at the higher part of the terahertz spectrum. As a result, the output power of the photomixer with the new nanogap electrode structure in the active region is two orders of magnitude higher than for a photomixer with typical interdigitated electrodes. The terahertz emission bandwidth also increases by a factor of more than two.

Turns out that, despite undergoing conventional peer review — so we assume — the conclusions were based on a mirage: namely, that the equipment Danner’s the group had used in their experiment functioned as they’d believed. Evidently, it didn’t.

According to the retraction notice:

In this Article, we presented a nanogap continuous-wave terahertz photomixer that exhibited a two orders of magnitude intensity enhancement compared with a photomixer with conventional interdigitated electrodes, which we used as a reference device. We estimated the powers of the devices based on the blackbody power of a Hg lamp, as described in the Methods section and shown in Fig. 3b of this Article. After receiving a communication from J. Mangeney, R. Colombelli, E. Peytavit, J. F. Lampin, M. Jarrahi, S. Barbieri, M. Wanke and M. A. Belkin, we measured the absolute output of our device using a loaned pyrodetector (Gentec-EO, THZ1.5B-BL-USB). We measured the absolute output power to be about 1 μW, instead of 100 μW as initially estimated and shown in Fig. 3b. Due to this error, we wish to retract the Article, even though the reported relative enhancement and the idea of using nanogap electrodes are still valid. We apologize to the readers for any adverse consequences that may have resulted from the paper’s publication.

The paper has been cited eight times, according to Thomson Scientific’s Web of Knowledge.

Danner, in an email, said:

Obviously we are all disappointed to retract a paper from Nature Photonics.

But he referred questions about the specifics of the problem to a co-author, who hasn’t responded to our queries.

Update, 4/9/13: Please see an update to this post, which explains the strikethroughs above.