Improving the Detection Limit in a Capillary Raman System for In Situ Gas Analysis by Means of Fluorescence Reduction.

Raman spectroscopy for low-pressure or trace gas analysis is rather challenging, in particular in process control applications requiring trace detection and real-time response; in general, enhancement techniques are required. One possible enhancement approach which enjoys increasing popularity makes use of an internally-reflective capillary as the gas cell. However, in the majority of cases, such capillary systems were often limited in their achievable sensitivity by a significant fluorescence background, which is generated as a consequence of interactions between the laser light and optical glass components in the setup. In order to understand and counteract these problems we have investigated a range of fluorescence-reducing measures, including the rearrangement of optical elements, and the replacement of glass components--including the capillary itself--by metal alternatives. These studies now have led to a capillary setup in which fluorescence is practically eliminated and substantial signal enhancement over standard Raman setups is achieved. With this improved (prototype) setup, detection limits of well below 1 mbar could be obtained in sub-second acquisition times, demonstrating the potential of capillary Raman spectroscopy for real-time, in situ gas sensing and process control applications, down to trace level concentrations.

Relative Intensity Correction of Raman Systems with National Institute of Standards and Technology Standard Reference Material 2242 in 90°-Scattering Geometry.

The U.S. National Institute of Standards and Technology (NIST) has certified a set of Standard Reference Materials (SRMs) that can be used to accurately determine the spectral sensitivity of Raman spectrometers. These solid-state reference sources offer benefits such as exact reproduction of Raman sampling geometry, simple implementation, and long-term stability. However, a serious drawback of these SRMs is that they are certified only in the backscattering (180°) configuration. In this study, we investigated if and how SRM 2242 (applicable for 532 nm) can be employed in a 90°-scattering geometry Raman system. We found that the measurement procedure needs to be modified to comply with the certified uncertainty provided by NIST. This requires a change in the SRM illumination that is possible only if we polish the side surfaces. In addition, we need to account for the polarization configuration of the Raman system by choosing the appropriate polarization of the excitation beam. On top of that, the spatial inhomogeneity of the luminescence light needs to be taken into account, as well as its behavior while traveling through the SRM bulk. Finally, we show in a round-robin test that the resulting uncertainty for the quantification of Raman spectra using the modified technique is on the order of ±1.5 percentage points.

In-line calibration of Raman systems for analysis of gas mixtures of hydrogen isotopologues with sub-percent accuracy.

Highly accurate, in-line, and real-time composition measurements of gases are mandatory in many processing applications. The quantitative analysis of mixtures of hydrogen isotopologues (H2, D2, T2, HD, HT, and DT) is of high importance in such fields as DT fusion, neutrino mass measurements using tritium ß-decay or photonuclear experiments where HD targets are used. Raman spectroscopy is a favorable method for these tasks. In this publication we present a method for the in-line calibration of Raman systems for the nonradioactive hydrogen isotopologues. It is based on precise volumetric gas mixing of the homonuclear species H2/D2 and a controlled catalytic production of the heteronuclear species HD. Systematic effects like spurious exchange reactions with wall materials and others are considered with care during the procedure. A detailed discussion of statistical and systematic uncertainties is presented which finally yields a calibration accuracy of better than 0.4%.