Volume 54, Number 9, 2000 APPLIED SPECTROSCOPY 1275 0003-7028 / 00 / 5409-1275$2.00 / 0 q 2000 Society for Applied Spectroscopy Steel Analysis with Laser-Induced Breakdown Spectrometry in the Vacuum Ultraviolet V. STURM, L. PETER, and R. NOLL * Fraunhofer Institut fu¨r Lasertechnik (ILT), Steinbachstraße 15, D-52074 Aachen, Germany Laser-induced breakdown spectrometry (LIBS) with multiple pulse excitation has been applied for the multielemental analysis of steel samples in the vacuum ultraviolet. The emission of the induced plas- ma was coupled into a Paschen–Runge spectrometer equipped with photomultipliers linked to high-speed multichannel signal electron- ics. Time-resolved signal evaluation yields a signi cantly improved signal-to-noise ratio for the plasma emission periods after a multi- pulse excitation. Reference materials for low-alloy steel grades were used to calibrate the measurements. The investigations concentrated on the light elements phosphorus, sulfur, and carbon using emission wavelengths in the range from 178.28 nm to 193.09 nm. For the rst time, limits of detection below 10 m g/g were achieved for the light elements phosphorus, sulfur, and carbon using LIBS. With these results the basis is established for future on-line applications of LIBS in the steel industry. Index Headings: Laser-induced breakdown spectrometry; Steel; Vacuum ultraviolet, Paschen–Runge spectrometer; Multi-pulse ex- citation, Limits of detection. INTRODUCTION The chemical analysis of steel is a demanding task for process control in steel making and for quality assess- ment of pre-products. The dominant analytical tools ap- plied nowadays are based on physical methods such as spark discharge-optical emission spectrometry (SD- OES), X-ray uorescence (XRF), or absorption spectros- copy of analytes in ames. These techniques are mostly used in off-site laboratories, and they require prepara- tional steps to present the steel sample in a physical state necessary for the respective method. The sample prepa- ration includes steps such as cutting, milling, grinding, or drilling, which are time consuming and require high maintenance efforts even in automated container labora- tories. Hence, there is an increasingly strong need to sim- plify or even avoid those steps of chemical analysis such as sample taking, sample transport, and sample prepara- tion. Laser-based analytical methods will play a key role for the development of on-line methods in metallurgical processing due to the following features: noncontact mea- surement at distances of centimeters to meters, high mea- suring speed, and sample preparation or conditioning by the laser beam itself. For an on-line analytical method, the multielement ca- pability is an important issue. Laser-induced breakdown spectrometry (LIBS) is a method enabling the simulta- neous analysis of various elements in a short time. How- ever, the analytical performance in terms of limits of de- tection reported for steel constituents so far does not achieve those of conventional methods such as SD- Received 8 December 1999; accepted 28 April 2000. * Author to whom correspondence should be sent. OES. 1–3 Hence the main goal of our work presented in the following was to improve the plasma excitation by multiple laser pulses and to make use of emission lines in the vacuum ultraviolet (VUV) spectral range. The mo- tivation of this work was to establish the basis for future on-line applications of laser-induced breakdown spec- trometry in the steel industry requiring limits of detection for light elements such as carbon, phosphorus, and sulfur in the range below 10 m g/g. EXPERIMENTAL Figure 1 shows the experimental setup. A Q-switched Nd:YAG laser (Continuum, Model Surelite I) operating at 1064 nm was used to excite the plasma. The Q-switch electronics of this laser were modi ed to generate up to three separated laser pulses within a single ashlamp pulse instead of a single laser pulse. The ashlamp rep- etition rate is 10 Hz. The in uence of multiple laser puls- es on the dynamics and physical state of the induced plasma was described elsewhere. 4 The laser emits three equal-energy pulses of ; 16 ns duration [full width at half-maximum (FWHM)] and with a time separation of 25 m s between the rst and the second pulse, and 40 m s between the second and the third, respectively. The burst energy amounts to 300 mJ. The laser beam is guided by two mirrors to a plano-convex lens with a focal length of 80 mm. The converging laser radiation is then guided via a mirror and an optical window into a measurement chamber. This chamber has a cover plate on top with a circular opening having a diameter of ; 12 mm. The sam- ple to be analyzed is put onto this cover plate and located in such a way that the opening is closed. The focus of the laser beam was adjusted so that the focal plane lies inside the sample with a distance of 6 mm from the sam- ple surface. The diameter of the craters formed on the sample surface is about 300 m m. The optical axis of the VUV spectrometer describes the line between the center of the diffraction grating and the middle of the entrance slit. This optical axis intersects the propagation axis of the laser beam on the surface of the sample. The angle between the surface of the sample and the optical axis of the spectrometer amounts to ; 188 . With this orientation it is ensured that radiation from the whole laser-induced plasma can be received by the spec- trometer aperture. The chamber has an internal volume of about 42 cm 3 . This volume is ushed with argon 5.0 at 1 atm and a ow rate of about 8 L per minute. The entrance window of the spectrometer is made of VUV transparent mag- nesium uoride. The spectrometer has a Rowland circle diameter of 750 mm and is equipped with a holographic diffraction grat-