Miniaturized Quantum Cascade Lasers Enable Fast Mid-IR Spectroscopy

Quantum cascade lasers have been gaining increasing attention as their capabilities are being demonstrated in a range of applications. One recent advance is the development of a miniaturized QCL, which when used as a light source, enables mid-infrared (mid-IR) scanning speeds much faster than those of conventional Fourier-transform IR (FT-IR). Ralf Ostendorf of the Fraunhofer Institute for Applied Solid State Physics (IAF) in Freiburg, Germany, recently spoke to Spectroscopy about this work.

In a recent paper (1), you presented information that showed how an external-cavity quantum cascade laser (EC-QCL) equipped with a custom made micro-opto-electro-mechanical systems (MOEMS) scanning grating in Littrow-configuration can be used in wavelength-dependent feedback element in a mid-IR EC-QCL, featuring a large scanning grating plate 5 mm in diameter. How does this technique improve on current techniques?

Initially, our goal was to miniaturize the laser source. We wanted to build a compact, matchbox-sized laser source and demonstrate the capability of laser-based mid-IR laser spectroscopy in different applications. We started a collaboration with our colleagues from the Fraunhofer Institute for Photonic Microsystems (IPMS) in Dresden, who are experts in developing MOEMS technology. Our idea was to use a small silicon chip with an implemented diffraction grating (so called MOEMS grating) as a wavelength selective element and combine it in an external cavity with a quantum cascade laser chip developed at Fraunhofer IAF. Both chips are pretty small, just a few millimeters, so we saw an enormous potential to miniaturize the laser source. The MOEMS grating is tilted by electrostatic forces and is driven in a resonant way, meaning it is tilting continuously with frequencies of up to several kHz. At this point we realized that we could also perform very fast spectroscopy even in real-time.

The innovation in this MOEMS EC-QCL, apart from the fact that it is no bigger than a matchbox, is that it is able to tune the wavelength over a broad spectral range of more than 300 cm^-1 in the mid-IR within just one millisecond. Using this laser as a light source for mid-IR fingerprint spectroscopy allows for recording of up to 2000 spectra per second (tilting the grating up and down during one period equals two complete wavelength scans) which is far faster than any conventional mid-IR spectroscopy, such as Fourier transform IR (FT-IR) spectroscopy. Thus, the MOEMS QCL opens the possibility of identifying and quantifying chemical substances in real-time.

Moreover, due to the spectral brightness of the laser, you can even measure aqueous solutions. Usually, IR light is strongly absorbed by water. If you use common techniques like FT-IR, you are able to measure small water films no thicker than 10 µm. We were able to identify caffeine dissolved in water (25 mg/L) through a 150-µm water film, and there is the potential to go to even thicker water films. This is very attractive for sensing solutions that are operated in a bypass flow cell.

Why is the tuning speed of an EC-QCL important for applications such as real-time monitoring and in-line process analysis?

In the end it comes down to the simple equation “time = money.” Today, if you want to analyze the chemical composition of a large amount of bulk material, such as pills in pharmaceutical production or nuts in a food processing plant, you need to take random samples, bring them to a lab and perform time-consuming chemical analysis using, for example, FT-IR spectroscopy. The high tuning speed of the MOEMS EC-QCL allows for the development of mid-IR sensing systems that can record 2000 spectra per second. This means in principle you can analyze 2000 nuts or pills per second, allowing for 100% control of your batch during the production process, without the need to take random samples. In this way you can speed up your production while having more control over your process at the same time.

Furthermore, due the high tuning speed of the MOEMS EC-QCL and the fact that the laser is emitting in the mid-IR fingerprint region, where the absorption lines of chemical substances are so characteristic and strong like nowhere else, you can trace a chemical reaction in real-time and see how much of your chemical reactants are already consumed and transformed into the final chemical product. In this way you can optimize chemical reactions on-the-fly, for example, to enhance yield or to speed up production cycles.

You have applied this technique to monitoring a Knoevenagel condensation reaction, which is a prototypical catalytic chemical reaction. What results have you seen so far?

The monitoring of the Knoevenagel condensation was performed by colleagues of mine at the Fraunhofer Institute for Chemical Technology (ICT) in Pfinztal. They still used a “slow” scanning EC-QCL; for example, recording one spectrum from 1050 cm^-1 to 1300 cm^-1 took about 3–4 seconds. The Knoevenagel condensation was chosen to act as a prototypical catalytic chemical reaction to demonstrate the capabilities of the EC-QCL based spectroscopy for chemical process monitoring. The spectra show that, despite a very low concentration of the involved reactants of 40 mM, very good signal intensity and low noise were achieved, both mostly resulting from the high energy density of the laser source. During the reaction time of about 15 minutes, a clear change in the absorption spectra could be observed. For example, a decreasing absorption band at 1160 cm^-1 attributed to the aldehyde group of the educt (4-methoxybenzaldehyde) and a rising absorption band at about 1180 cm^-1 attributed to the emerging C=C double bond of the product was monitored. This allowed for the application of chemometric quantification procedures and determination of the product composition during the whole time of the reaction. Based on these data the kinetics of the chemical reaction could be characterized and optimized for yield, conversion efficiency, and the influence of the involved catalyst.

Can you describe the results you have achieved in applying this technique to the standoff detection of explosives?

We are still using the “conventional” EC-QCL for the standoff detection of explosives, although we have plans to use our fast scanning MOEMS EC-QCL for this application as well in the near future. The standoff system developed at Fraunhofer IAF is based on a hyperspectral imaging technique. For example, the laser is used to actively illuminate scenery, and while tuning the wavelength of the laser, an IR camera system takes pictures from the diffusely backscattered light for each wavelength. From these data, spectra can be extracted using sophisticated algorithms to identify traces of chemical contaminations in the illuminated scenery, such as the remains of explosives. The group at Fraunhofer IAF developing this technique led by Dr. Frank Fuchs has already successfully demonstrated the reliable detection of small traces of explosives over distances of more than 20 meters. Using the MOEMS EC-QCL approach in the near future will enable real-time capability for this application as well.

What are your next steps in your research?

We are currently working on first real-time measurements spectroscopic measurements to demonstrate the capabilities of this technology as well as paving the way for first applications like a handheld sensor for hazardous substances, which is developed within the framework of a joint European Project called “Chequers.” We have already been able to show real-time absorption on a polysterene sheet and measure a complete absorption band of carbon monoxide over a spectral range of more than 200 cm^-1 in real-time.

We are also working further on miniaturizing the laser source and simplifying alignment procedures of the grating and micro-optics in order to increase the quantity of laser sources we can provide to potential end-users.

On the side of the quantum cascade laser chip, we are trying to widen the spectral tuning range of the laser to be able to address more absorption lines in spectroscopic applications. Regarding the MOEMS grating, we have some ideas with our colleagues at Fraunhofer IPMS to enhance the functionality of the laser source.

Reference

(1) J. Wagner, R. Ostendorf, J. Graham, A. Merten, S. Hugger, J.P. Jarvis, F. Fuchs, D. Boskovic, and H. Schenk, SPIE Photonics West 2015-Conference 9370.