The combination of a laser and time-of-flight mass spectrometry promises to enable semiconductor manufacturers to measure the size and type of individual particles in a vacuum chamber during the manufacturing process. Researchers at Lucent Technologies' Bell Labs led by chemist William Reents have developed a system that can identify a particle, its concentration in a gas flow and its size, all at the same time. A focused Nd:YAG laser is a key feature of a particle detection system developed at Bell Labs that can identify particles using ion-based time-of-flight mass spectrometry.Ion detectionThe experimental device uses an Nd:YAG laser to turn particles in a gas flow into atomic ions as they enter a vacuum chamber. An electric field then accelerates the ions. They leave the electric field and drift, eventually striking a microchannel plate, the standard detector for time-of-flight mass spectrometry. The ions with higher mass will have lower velocity and will take longer to reach the detector. Likewise, the ions with lower mass will move more quickly to the detector. "We monitor the ion signal as a function of time, and we can identify all of the atoms present from a single particle," Reents said. "From this data, we pull out information such as the identity of the element." The intensity of the signal tells the researchers how many ions of a particular element are present which, when combined with other information, tells them the identity of the original particle. The flow of gas is continuous, and the laser is constantly firing, so the rate at which the system detects a particle indicates its concentration in the gas. The group uses a high-intensity Nd:YAG to separate the particles into atomic ions. At the fundamental wavelength of 1064 nm, it uses 300-mJ pulses, and at the frequency- doubled wavelength of 532 nm, 150-mJ pulses. Reents said that the researchers have used lasers from Spectra-Physics Lasers Inc. of Mountain View, Calif., and Quantel SA of Les Ulis, France. Although the system is still in the experimental stage, Reents envisions developing it for monitoring semiconductor manufacturing in real time. An early application, he said, would be to use the system to look for "particle excursions," when the concentration of particles rises above a desired level. This occurrence indicates that the manufacturing machine needs cleaning, which can improve semiconductor quality control. He added that it also may prove useful in nanoparticle sizing and atmospheric particle monitoring.