Portland, Ore. - Less than two years after Lucent Technologies Bell Labs pioneered the quantum cascade laser and predicted more-sensitive spectroscopy, Georgia Tech researchers say they've validated the concept of single-chip spectra analysis. By integrating all the components in one device, such a spectrometer could enable lab-on-a-chip applications reminiscent of the handheld "tricorder" popularized by Star Trek.
No one has yet integrated all the components onto a single chip, but Boris Mizaikoff, an associate professor at the Georgia Institute of Technology, said he has proven the concept using available laboratory quantum cascade lasers, waveguides and detectors.
Now he is testing the device as an ultrasensitive spectrometer. By using the midinfrared signatures for molecules that have the characteristics of known explosives, toxins and other agents of interest, Mizaikoff has shown that a single chip can identify through spectroscopy almost any substance of interest after "sniffing" scant parts per billion of that substance. Ultimately, the single-chip lab would be housed in a handheld device.
In his initial tests, Mizaikoff coupled a hollow waveguide to a frequency-matched quantum cascade laser to irradiate a 1-milliliter gas sample. Conventional spectroscopy, by contrast, samples hundreds of milliliters. Yet sensing inside the photonic-bandgap material enabled the detection of levels down to 30 parts per billion.
"We are using very advanced laser technology to build chemical-sensing systems," said Mizaikoff. "In this case we are using quantum cascade lasers as semiconductor light sources. The beauty is that they emit in the midinfrared range, which is particularly useful because almost any organic molecule has a very specific vibrational pattern in the frequency range from about 3 to 20 micrometers [microns]." Mizaikoff was assisted by doctoral candidate Christy Charlton.
The waveguides channel the laser's light so that it illuminates only a very small amount of the gas or liquid being tested with a single wavelength of light, which corresponds to the fingerprint frequency of the substance to be detected. The waveguide for the liquid sensor was fabricated by Tel Aviv University (Israel) and the waveguide for the gas sensor was developed at OmniGuide Inc. (Cambridge, Mass.).
Quantum cascade lasers (see www.eetimes.com/issue/tech/showArticle.jhtml? article I D = 18310102) are similar in geometry to vertical-cavity surface-emitting lasers, but can be integrated into arrays on semiconductor chips for sensing spectroscopy. Several labs, such as professor Federico Capasso's group at Harvard University and Claire Gmachl's group at Princeton University, are working toward photonic crystals integrated with vertically emitting quantum cascade lasers. But the laser used in Mizaikoff's experiment was a quantum cascade laser that emitted from a cleaved facet perpendicular to the layered semiconductor structure. It was provided by Mizaikoff's collaborator in the experiment, Alpes Lasers SA (see www.alpeslasers.com).
When an electric current flows through a quantum cascade laser, electrons cascade down an energy staircase, emitting a photon at each step. Quantum-well structures create a powerful pulsed cascade effect in which many photons are generated from each electron. Instead of moving smoothly between levels, electrons tunnel from one layer to the next. As they do so, they jump from one energy level to another and ideally emit a photon with each jump.
In conventional lasers, an electron emits only a single photon by jumping from the semiconductor's conduction band to its valence band, where it is neutralized. But a quantum cascade laser arranges from 25 up to 75 quantum wells at slightly lower energy levels relative to each other, thus producing the cascade effect that allows as many as 75 photons to be created per electron. The quantum cascade laser is fabricated using molecular-beam epitaxy, a layer of atoms at a time, from alternating layers of gallium and aluminum. Each layer is slightly thinner than the one before it.
A waveguide attached to the output of the quantum cascade laser passes a single frequency of midinfrared light through a test sample. Then a detector measures the amount that was absorbed, thereby quantifying how much of a given substance is present. Mizaikoff's group demonstrated both a gas and a liquid sensor that used the same lasers and detectors, but different waveguides.
"We have two different sensing concepts, but they are utilizing a similar light source and detection technology," Mizaikoff said. "The main difference between the gas-phase and liquid-phase sensor is the waveguide. For the gas phase we use a hollow waveguide made from photonic-bandgap material-basically a hollow core fiber where the walls are made from a photonic-bandgap material."
By using the inside of a hollow fiber core that is only 700 microns in diameter, the entire hollow core could also serve as the 1-milliliter chamber for the gas to be tested.
For the liquid-phase sensor, Mizaikoff used a solid-core planar waveguide inserted into the liquid with the detector at the end of the waveguide. The laser "leaks" into the surrounding sample, enabling absorption spectroscopy to be performed on the signal at the detector.
Quantum cascade lasers require very little silicon real estate, since they only need eight wavelengths to begin lasing. A 20-micron laser, for instance, could be as compact as 160 microns.
By arraying many different frequency lasers on a chip, spectroscopy could be performed on a variety of benchmark substances simultaneously. Measuring such organic building blocks as hydrogen and carbon to determine the amount of each could identify a wide variety of ad hoc substances.
"The laser itself is only a few hundred microns, so you could have an array of them illuminating a very, very small sample volume," said Mizaikoff.
Detectors small enough to be integrated onto a chip with the laser will also have to use the quantum effect to achieve matching-sized detectors. For instance, quantum-well photo-conductive devices can be made as small as the quantum cascade lasers. And various types of thermal detectors can also be downsized to be "as small as a pinhead," Mizaikoff said.
Integration of all the components-quantum cascade laser, photonic crystal waveguide and quantum-well photo detectors-might also make use of microelectromechanical systems (MEMS) to guide the test sample from the environment to the waveguide and to flush it after the test.
"Our next-generation devices might integrate MEMS components. The liquid-phase device will definitely have microfluidics," said Mizaikoff. "Currently, we are trying to hybridize the components to create a smaller platform. That's the first step, but the vision is to eventually fabricate all the components on the same wafer-you can fabricate the semiconductor laser, the waveguide and maybe even the detection devices all on one wafer."
In the immediate future, Mizaikoff plans to improve the sensitivity of his liquid-phase sensor by "another one or two orders of magnitude" (10- to 100-fold).
"We will have a new waveguide for liquid-phase sensing that will be a single-mode waveguide matching the single-mode laser," said Mizaikoff.
Today only gas-or liquid-chromatography can detect such fine amounts, but not by sniffing the sample. Instead, the suspect object must be swabbed first. Then the swab is inserted into a detector chamber. Mizaikoff's device, on the other hand, could be set up to monitor the air or water continuously.
"This technology leads to single-digit parts-per-billion-level quality measurement," said Mizaikoff. "We envision sensors that continuously measure water and air at parts-per-billion levels instantaneously and with molecular selectivity."