Most naturally emitted ultraviolet and visible radiation of flames, called chemiluminescence, is caused by short lived electronically excited intermediate species such as OH*, CH* or C2* formed during chemical reactions . An example is depicted in Fig. 1 (left panel) from a Bunsen burner, operated with a stoichiometric mixture of methane and air. If the radiation emitted by this flame is dispersed in a spectrometer into its wavelength components, spectra can be observed like those shown in Fig. 1 (right panel). On recognizes characteristic emission bands originating from the above mentioned species in their respective electronically excited (A) states of OH* (310 nm), CH* (388, 431 nm) und C2* (473 nm), respectively. It is therefore of interest to investigate if the intensities or spectral shapes of these band systems can provide quantitative information on, e.g., local fuel/air ratios, heat release rate (HRR) or chemical processes of the combustion system investigated. If so, this would be of great importance for technical combustion systems, since detecting chemiluminescence radiation is a cheap and non-intrusive method for monitoring the combustion event in environments such as power plants, waste incinerators or combustion engines.
Figure 1: left: chemiluminescence radiation from a Bunsen burner flame.
right: spectrum of the chemiluminescence radiation emitted from a premixed methane/air flame. Vibronic emission bands can be recognized from OH*-, CH*- and C2*-radicals.
From a technical viewpoint operating flames with chemiluminescence detection is a simple and low budget method for continuous monitoring and control of large scale combustion systems, such as in power plants or combustion engines. Therefore, basic research in quantifying chemiluminescence with respect to heat release rate and fuel/air ratio measurements is of high importance. For this purpose, in our group we investigate stable, laminar counterflow diffusion as well as turbulent swirl flames at atrmospheric pressure, respectively, at a variety of operating conditions. In the first case flame stretch and fresh gas composition (fuel composition, fuel/air ratio) can be varied, whereas in the second case besides fresh gas composition the extent of swirl can be changed within certain limits. For generating the first type of flames a special burner was constructed and built as depicted in Fig. 2 (left panel). Chemiluminescence spectra for each species were recorded with the spectrometer/camera combination shown in the right panel of Fig. 2, and spatial profiles of chemiluminescence intensities were evaluated as a function of distance from the fuel supply cylinder (cf., Fig. 3).
Figure 2: counterflow diffusion burner: within the optically accessible combustion chamber an air (from below) and fuel (from above) flow impinge on each other and form a counterflow diffusion flame around the fuel cylinder (consisting of porous sintered bronze) as shown in the right cross section.
right: experimental arrangement for taking spatially resolved (perpendicular to the fuel cylinder surface) chemiluminescence spectra in the counterflow flame.
With respect to obtaining a measure for the heat release rate in flames chemiluminescence measurements will be compared with an alternative method, i.e., the detection of hydroxyl radicals (OH) and formaldehyde (H2CO) in their electronic ground states (X). The latter will be detected using the modern laser-spectroscopic diagnostic method of laser-induced fluorescence (LIF). By this technique laser beam, whose wavelength is absorbed by the respective molecule, is formed into a thin light sheet and sent through the flame gases. Absorbing molecules distributed within the excitation volume will be excited and subsequently send out fluorescence radiation in a characteristic wavelength region, which can be analyzed for species concentration (and temperature). In the specific application the product of the two-dimensional distributions of OH and H2CO species is proportional to the local heat release rate in the common overlap region of two separate light sheets used for each species excitation.This procedure for HRR measurement is based on the fact that the most important elementary reaction responsible for heat release in premixed flames is given by:
CH2O + OH --> HCO + H2O,
and that the product of the reactant concentrations is a measure for this quantity. Experimentally, this requires the simultaneous and instantaneous capture of a relative signal intensity from both species - if possible in a two dimensional field of view . Subsequently, this product of both LIF signal intensities, (ILIF, CH2O (T) × ILIF, OH (T)), is formed, which then represents the spatial region within the light sheet where the heat release takes place. Such results obtained from in-situ laser diagnostic measurements in flames are an important data base for, on the one hand the validation of the chemiluminescence measurements as well as chemical kinetics mechanisms for simulating these combustion systems.
 A. G. Gaydon and H. G. Wolfhard, Flames: Their Structure, Radiation, and Temperature (Chapman and Hall, London, 1978).  P. H. Paul and H. N. Najm, "Planar laser-induced fluorescence imaging of flame heat release," Proc. Combust. Inst. 27, 43-50 (1998).
Figure 3: spatial profile of chemiluminescence intensities of OH*, CH*, C2* und CO2*, as measured in the diffusion flames of fig. 2.