A cw laser absorption diagnostic for methyl radicals

A cw laser absorption diagnostic for methyl radicals
of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  J. Qumt. Spectrosc. R at. Tranrfer Vol. 49, No. 5, pp. 559-571, 1993 Printed in Great Britain. All rights reserved 0022-4073/93 $6.00 + 0.00 Copyright 0 1993 Pergamon Press Ltd A CW LASER ABSORPTION DIAGNOSTIC FOR METHYL RADICALS D. F. DAVIDSON,? A. Y. CHANG, M. D. DI ROSA, and R. K. HANSON High Temperature Gasdynamics Laboratory, Mechanical Engineering Department, Stanford University, Stanford, CA 94305, U.S.A. Received I April 1992; received for publication 20 October 1992) Abstract-Absorption of narrow-line laser radiation by methyl radicals produced at high temperatures was studied in the Herzberg B, band near 216nm. cw radiation for these measurements was generated in a ring dye laser system using intracavity BBO frequency doubling. Methyl radicals were produced by the shock wave heating of five selected source compounds: azomethane, methyl iodide, tetramethyl tin, tetramethyl silane and ethane. The variation with wavelength of the CH, absorption coefficient between 210 and 225 mn was measured at 1625 K and 1.25 atm using ethane/Ar and methyl iodide/Ar gas mixtures. The variation of the absorption coefficient with temperature from 1350 to 2450 K was measured at 216.615 mn using four source compounds. Using this wavelength for CH, measurements resulted in a typical detectivity limit of 2 ppm at 1650 K, 1 atm, L = 10 cm, with a SNR of unity and a detection bandwidth of 500 kHz. INTRODUCTION The methyl radical is an important intermediate species in the pyrolysis and oxidation of aliphatic hydrocarbons such as methane and ethane. Accurate concentration measurements of CH3 in these systems, however, have been limited by the lack of a sensitive, selective, fast time-response optical diagnostic and by the ever-present interference to detection from other species with spectral features in the same wavelength regime. Despite these circumstances, CH3 U.V. absorption techniques have been applied to the measurements of rate coefficients in kinetically-simple high temperature chemical systems,lA and occasionally to the analysis of complex combustion systems59 6 One of the first comprehensive shock tube studies of methyl reactions using absorption was conducted by Gliinzer et al.’ These authors studied the recombination of methyl radicals to form ethane using absorption of relatively broadband 216 nm radiation. This radiation was generated using a high pressure xenon arc lamp, dispersed using a quartz prism and a monochromator and detected with a photomultiplier. Miiller et al,’ Hwang et al3 and Tsuboi4 used similar methods in their shock tube studies of methyl reactions. CH, absorption measurements were first done in a low pressure flame by Harvey and Jessen,7 also using a Xe lamp system. Tsuboi’ studied methane oxidation in a shock tube with several optical diagnostics, including CH, absorption at 216 nm using a xenon lamp. Gardiner et al6 repeated a portion of this latter study using a zinc lamp which has an atomic resonance line at 214 nm. These earlier experimental methods have significant limitations. Interference absorption from combustion reactant or product species such as O2 and NO can complicate interpretation of the data, and significant absorption from certain species cannot be avoided with either the fixed narrow-linewidth Zn lamp system or with the tunable, but spectrally-broad, Xe lamp system. In addition, neither of these earlier systems takes full advantage of the larger magnitude of the peak of the CH3 absorption feature, nor do they provide the large spectral intensities generally associated with lasers. A spectrally narrower, tunable and more intense laser source would help overcome these limitations. tTo whom all correspondence should be addressed. 559  560 D. F. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFE  VIDSON t al Work in our laboratory has concentrated on the accurate and sensitive detection of combustion species by the use of cw ring dye laser absorption. Although narrow-line cw radiation at wavelengths as short as 144 nm has been generated through frequency tripling,’ cw power levels sufficient for use in transient experiments can only be obtained through either second-harmonic generation or sum-frequency mixing. Performing either technique intracavity leads to enhanced power levels.g Until recently, the shorter wavelength limit for intracavity doubling was 238 nm (achieved using LFM, Lithium Formate Monohydrate). ‘O The advent of a frequency-doubling technology based on /I-BaB,O, (BBO) has permitted access to wavelengths as short as 205 nm, and to even shorter wavelengths through frequency mixing. ‘I At wavelengths between 205 and 238 nm, strong absorption bands of several significant combustion species, including NO, 0, and CH,, appear. The use of this BBO laser system to detect NO and 0, has been described elsewhere.‘2*‘3 Here we describe the results of high-temperature shock tube/BBO laser system experiments which were performed to determine the optimal wavelength for a CHS absorption diagnostic and the temperature dependence of the absorption coefficient at this wavelength. A brief discussion of the effect of absorption by certain interfering species is also included. METHOD The apparatus, shown schematically in Fig. 1, consists of two primary components: the shock tube and the laser absorption system. High temperatures were generated behind incident and reflected shocks in a pressure-driven shock tube. The stainless steel shock tube driven section is 6.25 m in length with an i.d. of 14.3 cm. The driver section is 2.0 m long, with a 5.0 cm i.d., and has a diverging nozzle after the diaphragm/isolation-valve station. The vacuum system employs a turbomolecular pump and several zeolite-trapped mechanical pumps. The ultimate pressure of the system attained by pumping for 30 min between shocks is 8 x 10T6 orr; the combined leak and outgassing rate was less than 16 x 10V6 orr-min- ‘. Shocks were produced by bursting polycarbon- ate diaphragms (0.13 and 0.26 mm in thickness) against a rigid cutter. In all the experiments described, helium was used as a driver gas and 99.9995% argon (Matheson) was used as the driven carrier gas. Gas mixtures were prepared by partial pressures in a mixing tank with a large enough volume and at a great enough pressure to permit the use of the same gas mixture for six shock experiments. Mixing was accomplished with a magnetically-driven stirring vane. Temperature and pressure were calculated from the incident shock velocities using an ideal shock code. The shock Vacuum Pump Fig. 1. Schematic of the apparatus: WM-wavemeter; SI-scarming interferometer; Det.-photodiode detector; T.F. gauges-thin film gauges.  cw laser absorption diagnostic for methyl radicals 561 velocity was determined from shock arrival times measured using four platinum thin film gauges spaced at intervals of 600 mm. In separate experiments with this shock tube, the temperature, which was measured using the laser-scanning temperature monitor described in Chang et aLi4 and the pressure, which was measured using a piezoelectric transducer, were shown to vary from those values predicted using the local shock velocity by no more than the experimental uncertainty of 1.5% over the approx. l-2 msec of reflected shock test time available. The ring dye laser (Coherent model 699) was run with stilbene 3 dye (Lambda-Physik) and pumped with 5-6 W of all-lines U.V. from an Ar+ laser (Coherent model Innova 25/5). A custom BBO angle-tuned intracavity-doubling system was used to obtain approx. 1 mW of U.V. radiation near 216 nm (most experiments were done at a fixed wavelength of 216.615 nm), with a typical spectral width of less than 0.001 cm-‘. The crystal was cut for Brewster-angle incidence and for Type I phase-matching at 432 nm. With this system, we were able to produce useful levels of frequency-doubled output from 209 nm to beyond 220 mn. A portion of the visible (undoubled) laser output was passed into a wavemeter (Burleigh model WA-20) which measured the laser wavelength, and another portion was passed through a 2 GHz FSR scanning interferometer (Spectra-Physics model 470-2) to monitor laser mode quality. The U.V. beam was split into two approximately equal intensity components, one passing directly onto a detector to record the incident laser intensity (lo) as a function of time, and the second passing through uncoated fused silica windows, across the 14.3 cm dia of the shock tube 18 mm from the end wall and onto a second detector to provide the transmission measurement (I). Prior to each experiment, the detectors were optically balanced to obtain a null absorption signal (AZ = I0 - I) and the beams were centered on the detectors to minimize sensitivity of this null signal to shock tube vibration. The absorption signal (AZ) and the incident signal (lo) were recorded on a digital oscilloscope (Nicolet model 4094 with model 4562 vertical amplifiers) sampling at 0.5 psec per point. This dual-beam absorption method was efficient in rejecting common-mode noise leading to improved absorption detection limits relative to a one-beam absorption system. The detectors employed photodiodes (EG&G model WlOOBQ) and had electronic risetimes (I/e) of less than 0.5 /lsec. DATA REDUCTION AND KINETIC MODELLING Knowledge of the absorption coefficient combined with a radiation source of linewidth much narrower than the absorption feature width permits use of Beer’s law to directly interpret absorption data traces, i.e. ln(Z,/Z) = k&xcn,L, where I,, and Z are the incident and transmitted intensities at wavelength Iz, kn is the temperature- dependent absorption coefficient (atm-‘-cm-‘) at wavelength 1, PT is the total pressure (atm), xCHl is the mole fraction of the absorbing species CH,, and L is the absorber path length (cm). In order to determine the absorption coefficient of CH3, the absorption time histories were first converted to values of kA x xCH, vs time, and then were compared to CH3 mole fraction histories computed using the reaction mechanism of Table 1. The best-fit value of kA, for the temperature of the experiment, is extracted from the best-known portion of CH, time history, usually the period immediately after the arrival of the reflected shock. Only mixtures and shock conditions which were predicted to have peak yields of methyl radicals near the theoretical maximum (i.e., complete conversion of source species to CH,) were selected for the actual calibration in order to minimize the uncertainty associated with fitting the data with a kinetic mechanism. The reaction mechanism used is a subset of the H/C reactions found in the mechanisms of Miller and Bowman” and Glarborg et al,16 with the addition of several source species reactions described in the next section. Computations were done using the Sandia CHEMKIN-II program” and the 1991 Sandia Thermodynamic Database.” METHYL RADICAL SOURCES The selection of a suitable methyl radical source for use in shock tube experiments is affected by many factors including absorptivity, decomposition products and mechanism, and background  562 D. F. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFE  VIDSON t al Table 1. CH, reaction mechanism. - No. - 1 2 3 4 5 6 I 8 9 10 11 12 13 - uote: Reaction C&&=~CI-&+NZ CH31+h4=CH3+I+M SiC,H,2+M=Si+4CH3+M SnC,H,2=Sn+4CH3 C&+M=CH,+CH,+M &H~+H=CH,+CH, CH3+CH3=r$H,+H2 CH3+c H3=CHH,+CH4 CH3+H=CHz+H2 CH.,+H=CH3+H2 CJ%+H=C&+H2 w6+cJ53=c1 ~ CH3+M=CX&+H+M A 8.OElO 3.OE14 2.OE12 3.OE12 1.3E13 l.OE14 2.1E14 2.1E14 9.OE13 2.2EO4 5.4EO2 5.5E-1 1.9E16 B 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0’ 3.0 3.5 4.0 0.0 l- zyxwvutsrqponmlkjihgfedcbaZYXWVUT E Ref. 26cKlO. * 32OcKl. * 14000. * 47000. * 78400. * 0. 15 19200. 16 19200. 15 15100. 15 8750. 15 5210. 15 8300. 15 91600. 16 Rate coefficients are given in the form A ? exp(-E,/RT) (mol-cm~3-s~‘); ZA calories-mol.‘). * see the text for details. absorption from source species or subsequent reactant products. Many possible CH, source species are liquids at room temperature and may absorb onto the walls of the mixing vessel and shock tube in unknown amounts. Because of this absorption, it was necessary to compare the methyl yields of several source species to ensure that known reproducible quantities of methyl radicals are formed. In the following section, the results from experiments with azomethane, methyl iodide, tetramethyl tin, tetramethyl silane and ethane are compared. Azomethane (CH,N:NCH,) was srcinally chosen because it cleanly decomposes into two methyl radicals and N,. This is advantageous since, at the conditions of the present experiments, Nz plays no role in the chemistry. Azomethane is not commercially available and was synthesized using the method of Jahn.” This method is cleaner and simpler than the more frequently referenced method of Renaud and Leitch.” It also has the advantage of involving the production of a relatively stable intermediate which can be stored. Azomethane produced using the Jahn method may have trace water, ethane, CuCl, or acetal-based impurities. When azomethane is used in highly diluted (50-100 ppm) mixtures with argon, calculated yields of methyl radicals in the reflected shock regime can reach 97% or greater (of the theoretical maximum yield of 2 methyls per azomethane molecule). At very low shock temperatures (below 1250 K), where there is an extended CH, formation time, the peak yields are reduced and are determined by the rate for the azomethane decomposition, reaction 1, and the rates for reactions 6, 7 and 8 (see Table I). At temperatures higher than this level, the absorption coefficient derived from comparing the absorption data to the detailed kinetic model predictions is insensitive to the actual rate coefficients, though improved long-time fit can be achieved by adjusting the rate coefficients for reactions 7 and 8. The rate coefficient listed for reaction 1 is 10 x larger than that reported by Miiller et al, and gives improved fits to the lowest temperature data. An example data trace of the measured CH3 mole fraction vs time in an acceptable azomethane decomposition reflected shock experiment is shown in Fig. 2, along with calculated profile. In Figs. 2-6 the diagnostic wavelength used was 216.615 nm. The modelled peak CH, mole fraction is 94ppm, which represents a conversion yield of 98%; the estimated uncertainty in this calculated yield is +2%/- 10%. The time response of this trace is limited by the shock transit time across the probe laser beam, typically 1 psec. The noise, using the common-mode rejection scheme described earlier, is typically 0.2% of the lo signal. This corresponds to a CH3 detectivity limit of approx. 2 ppm. The spike near - 75 psec is a result of beam steering by the incident shock. The sharp absorption peak at time zero is a result of the combination of the slight absorption by azomethane, absorption from rapid methyl formation and beam steering by the reflected shock. The absorption fraction corresponding to 94ppm CH, in this trace is 17%; this was the mole  cw laser absorption diagnostic for methyl radicals 563 1445 K, 1.24 tm 48 ppm CHaN:NCHdArgon Fig. 2. Azomethane decomposition. The reflected shock conditions are 48 ppm azomethane/Ar, 1455 K, 1.24 tm; (---) kinetic model; IcL = 110 atm-l-cm-‘. fraction used to infer zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA n. Figure 3 shows the data of Fig. 2 plotted as l/x,-.,, vs time. The linearity of this plot supports the view that the decay is dominated, in the first 200 psec, by a second-order loss mechanism consisting of reactions 6,7 and 8. This is consistent with the calculated contribution factor analysis. Azomethane has limited use as an instantaneous methyl radical source at high reflected shock temperatures experiments (> 1900 K) as it first decomposes rapidly in the accompanying incident shock region. Under these extreme conditions, incident shock temperatures are sutlicient to decompose the azomethane into two methyl radicals, and these radicals have sufficient time (> 150 psec particle time at 0.2 atm 1000 K) to recombine into ethane. Then, behind the reflected shock, this ethane acts as a slower-releasing reservoir for methyl radicals and the CH3 profile appears to have a plateau at early times, This effect was noticed first by Glanzer et al.’ Methyl iodide (CH,I) was considered as a source species because it decomposes rapidly above 1600 K, and the iodine-related chemistry was expected to be insignificant in the temperature and pressure regime of the present experiments. 2’ Methyl iodide does not absorb on the mixing vessel and shock tube walls as much as the other sources tested. Adsorption, which usually manifests itself as a lower source species concentration in the first shock of a gas mixture, was not evident in the 6 iZ E 6 i? z YZ 4 1445 1.24atm , 46 ppm CH3N:NCH3/Argon 2 Fig. 3. Example data reduction. The data from Fig. 2 are plotted as l/xcH, vs time; (---) least-squares fit to data.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks