The CO2 laser, commonly recognized as an industrial workhorse, is not high on the list of potential optical pumping devices, at least not for quantum-well unipolar lasers. But optical pumping a new semiconductor laser is exactly the function of a pulsed transversely excited atmospheric CO2 laser used by Francois Julien and Olivier Gauthier-Lafaye of the University Paris-South.Optical pumping eliminates the need for doped cladding and metallic contacts common in other unipolar semiconductor lasers and enables the quantum fountain laser to avoid large free-carrier absorption losses at wavelengths above 10 µm.Julien's new unipolar device, termed a quantum fountain laser, emits at long IR wavelengths. His current prototype achieves peak pulse powers of 3 W per facet, which he said is a record value for semiconductor lasers emitting at 14.5 µm. "Our goal is to achieve much larger output powers (>100 W) by optimizing the pump beam coupling in the quantum fountain laser," Julien said. He attributed his prototype's current low output to its poor 2.5 percent pump coupling efficiency; that is, 97.5 percent of the pump energy never gets coupled into the quantum fountain waveguide. His pump source is a miniature transversely excited atmospheric laser from Edinburgh Instruments Ltd. in Edinburgh, UK. "The name 'fountain' comes from the fact that electrons in the quantum well are circulated between the three states," Julien explained. His prototype shares some features of newer quantum cascade lasers, particularly a groundbreaking device that lases in GaAs material recently introduced by researchers at Central Laboratories of Thomson-CSF in Orsay. Julien's laser departs from the quantum cascade fold in that it incorporates an optical pump, which circumvents the doped layers and metallic contacts on which quantum cascade lasers rely for current injection. The quantum fountain design thus avoids large losses of free-carrier absorption at wavelengths above 10 µm. "These internal losses are much smaller in our [quantum fountain] laser because we don't need doped cladding layers and metal contacts," Julien said. "Therefore, [these] lasers are expected to work better at wavelengths above 10 µm." Carlo Sirtori, leader of quantum cascade laser research at Thomson-CSF, agrees. "Normally, having one laser pump another laser is inconvenient. But when going for a very long wavelength, every contact is a source of loss," he said. One way to reach longer wavelengths, he added, is to optically pump. The Thomson-CSF electrically pumped quantum cascade laser currently emits at 9.4 µm, although Sirtori has demonstrated that wavelengths as high as 11.5 µm are possible with his device. Lucent Technologies' Bell Labs in Murray Hill, N.J., recently was able to reach 17-µm wavelengths with its quantum cascade technology. Julien's quantum fountain laser functions through electron transitions between three confined states in the conduction band of a semiconductor quantum well -- GaAs with AlGaAs barriers. The transversely excited atmospheric laser's beam is focused onto the quantum fountain laser with a 3 × 0.25-mm spot using cylindrical optics. Its energy promotes electrons from the ground state to the second excited state; lasing transition occurs between the second and first excited state.