Mid-infrared (MIR) light is rich with molecular "fingerprint" information that can be used to detect substances from atmospheric pollutants to cancer cells. While some lasers already operate in this region, enabling a variety of spectroscopy applications, their linewidth is relatively narrow, which limits the types of substances they can detect at any given moment. Now a team of researchers from Germany and Spain has developed a laser system with phase-coherent emission from 6.8 to 16.4 μm and output power of 0.1 W. That is broad and powerful enough, they said, to detect subtle signs of cancer early in its development. Molecules absorb portions of the MIR spectrum in ways that are unique to their atomic structures, and their absorption patterns provide a means of identifying the molecules with great specificity, even in low concentrations. The emission spectrum of the laser and corresponding molecular fingerprint regions. Courtesy of the Institute of Photonic Sciences (ICFO). "Cancer causes subtle modification in protein structure and content within a cell," said professor Dr. Jens Biegert, a group leader at the Institute of Photonic Sciences (ICFO) in Barcelona. "Looking at only a few nanometer range, the probability of detection is extremely low. But comparing many of such intervals, one can have an extremely high confidence level." The new laser system generates MIR pulses via difference-frequency generation driven by the nonlinearly compressed pulses of a Kerr-lens mode-locked Yb:YAG thin-disc oscillator. It features a repetition rate of 100 MHz and pulse durations of 66 fs — so short that the electric field oscillates only twice per pulse. Staff scientist Dr. Ioachim Pupeza (left) and postdoctoral researcher Oleg Pronin helped develop a laser system that emits ultrashort pulses of mid-infrared light. These pulses can be used to detect trace molecules in gaseous and liquid media. Courtesy of Thorsten Naeser/Ludwig Maximilian University. "Since we now possess a compact source of high-intensity and coherent infrared light, we have a tool that can serve as an extremely sensitive sensor for the detection of molecules, and is suitable for serial production," said project leader Dr. Ioachim Pupeza, a staff scientist at Ludwig Maximilian University of Munich (LMU). The LMU and ICFO researchers aim to use their MIR laser to identify and quantify disease markers in exhaled air. Many diseases, including some types of cancer, are thought to produce specific molecules that end up in the air expelled from the lungs. "We assume that exhaled breath contains well over 1000 different molecular species," said Dr. Alexander Apolonskiy, an LMU group leader. However, the amount of molecular biomarkers present in exhaled breath is extraordinarily low, meaning a diagnostic tool would need to be capable of detecting concentrations of at least one part per billion. The next step will be to couple the new laser system with a novel amplifier that would increase its brightness and boost sensitivity one part per trillion. Detecting MIR signatures The laser's output spans more than one octave. Until now, the researchers said, such broadband emission has only been available from large-scale synchrotron sources. Other more compact MIR sources, such as quantum cascade lasers (QCLs), have narrower linewidths. Tuning them to different sensing bands is time consuming, and combining multiple QCLs emitting in different parts of the MIR would be cost-prohibitive, Biegert said. Meanwhile, the laser system's 100-MHz pulse train is hundreds to thousands of times more powerful than state-of-the-art frequency combs that emit in the same range, the researchers said. Detecting broadband MIR signals presents its own problems, however. Detectors for this region have poor signal-to-noise ratios unless cooled with liquid nitrogen, the researchers said. In this case, electro-optical sampling proved to be a better option. Well-established for the terahertz range, the technique is less common in the fingerprint region. "In the MIR range, there are not many groups who have implemented this already, because you need a broadband, phase-stable MIR pulse and an ultrashort sample pulse at the same time, which is quite challenging," Pupeza said. Having solved that problem with their broadband laser, the team now could use electro-optical sampling to extract the data they wanted. In a nutshell, the process works like this: The electric field of an MIR pulse alters the birefringence of a crystal. This change can be measured by observing how the polarization of slightly shorter near-infrared (NIR) pulse is changed while propagating through the same crystal at the same time. In the end, only the NIR pulse is measured directly. "Therefore, one big advantage is low-noise detection in the NIR, even though one obtains information on spectral components in the MIR," said Ioachim Pupeza. "You only need to perform a Fourier transform numerically to get the spectrum of the pulse once you have its electric field." The research was published in Nature Photonics (doi: 10.1038/nphoton.2015.179).