Document NeY2b4G6DNNVyEbRQ9yj9KOXb

CNRS-LCSR-Orlans 16p. AR226-1944 EXTENDED LABORATORY STUDY OF THE ATMOSPHERIC DEGRADATION OF FLUORINATED ALCOHOLS Final Report of Amendment N1 to Subcontract N3529 between CNRS and RAND Corporation (TRP: Telomer Research Programme) February 2005 1 FOREWORD This report covers work from July 2004 to December 2004 and contains new information from work carried out at Valencia, Wuppertal and Orlans. It also contains plans for future work in 2005. This work has been carried out by Graldine Solignac, Wahid Mellouki and Georges Le Bras (CNRS-LCSR) Orlans, with the valuable contributions of Dr Ian Barnes (University of Wuppertal), Dr Klaus Wirtz (CEAM Valencia), Dr Howard Sidebottom (University College Dublin) and Dr Robert Waterland (DuPont, Wilmington, Delaware). 2 SUMMARY The main goal of the project was to determine the possible photolysis rate of the aldehyde, C6F13CHO, produced in the OH-initiated oxidation of the fluoroalcohol, C6F13CH2CH2OH. The related environmental issue is that the photolysis of this aldehyde under atmospheric conditions may compete with its OH reaction, hence reduce or even suppress production of the carboxylic acid, C6F13C(O)OH, through the OH reaction under low NOx conditions. The photolysis experiments under solar radiation were planned at the outdoor European reactor EUPHORE. Two sources of the aldehyde were initially proposed: the OH-initiated oxidation of C6F13CH2CH2OH and the OH-initiated oxidation of C6F13CH=CH2, both in the presence of air. To overcome the difficulty of measuring accurate photolysis rates, J, of the aldehyde by monitoring its consumption because of it additional loss by dilution, J was proposed to be measured by monitoring the production rate of CF2O (the photolysis of C6F13CHO is expected to produce 6 molecules of CF2O). EUPHORE experiments have been carried out using first the OH + C6F13CH2CH2OH reaction to produce the aldehyde, since simulation calculations have shown that the OH + C6F13CH=CH2 would have induced more undesirable secondary chemistry. Experiments have been extended to the OH + C8F17CH2CH2OH system to produce the corresponding aldehyde, C8F17CHO. Those experiments were not conclusive regarding the chemistry of the aldehydes C6F13CHO and C8F17CHO, due in particular to interferences in infra red spectra of fluorinated reactant and products (the consumption of the initial alcohol could not exceed 20%). In order to try to overcome the difficulties encountered with the above system, experiments have been carried out using Cl-initiated oxidation of C6F13CH2OH in air to produce directly C6F13CHO. However, infra red spectra were still too complicated to clearly distinguish the spectrum of the aldehyde. CF2O was detected as a reaction product but it has not been possible to derive reliable data on the loss rate of the aldehyde from simulation calculation of the concentration-time profiles of CF2O. 3 Additional experiments have been carried out on the Cl-initiated oxidation of C6F13CH2OH in air using the glass photoreactor of 480L of the University of Wuppertal. The aim was to determine accurately the infra-red spectrum of C6F13CHO in support to future experiments at EUPHORE. In addition, the rate constants of the reactions of Cl atom with C6F13CH2OH and C6F13CHO have been measured at room temperature. These data were needed to simulate the EUPHORE experiments carried out on the Cl-initiated oxidation of C6F13CH2OH. Other experiments are planned in the near future at EUPHORE to look directly at the photolysis of shorter aldehydes (C3F7CHO and C4F9CHO) which are available commercially through their corresponding hydrates. Few tests have been already conducted on C3F7CHO using an indoor smog chamber. RESULTS 1. OH-initiated oxidation of C6F13CH2CH2OH and C8F17CH2CH2OH at EUPHORE 1.a. OH-initited oxidation of C6F13CH2CH2OH (June 28th ) The experiment was performed using the photolysis of HONO as OH radical source. Initial concentration of C6F13CH2CH2OH was 250 ppbv. The loss of C6F13CH2CH2OH in absence of light was found to be comparable to that due to dilution (corresponding to the loss of SF6) indicating that the wall loss of C6F13CH2CH2OH is negligible in our experimental conditions. During the experiment, the consumption of C6F13CH2CH2OH due to its reaction with OH radicals was estimated to be around 16 %. GC-MS analysis showed two main products of the reaction. These products could not be not positively identified, however, one of them could be the aldehyde (C6F13CH2CHO). The FTIR analysis was rapidly complicated following the formation of the reaction products and the overlap of the fluorinated compounds with similar IR spectra. However, the IR spectra obtained during the experiments showed the presence of an unidentified band 1194-1264 cm-1 which could not be attributed to CF2O nor to C6F13C(O)H.. Figures 1-7 show examples of the IR spectra obtained during the experiments and the concentration-time profiles of C6F13CH2CH2OH integrated on band 1308-1097 cm-1 and of an unknown reaction product (integrated on band 1194-1264 cm-1 on residual spectrum with no further treatment) 4 It is important to note that in absence of spectra of pure sample of C6F13CH2CHO and C6F13CHO, it was not possible to quantify the formation of these two aldehydes during the experiments. Absorbance 1100 1150 1200 1250 Wavenumber (cm-1) 1300 1350 Figure 1: Residual spectrum after subtraction of C6F13CH2CH2OH Absorbance residual spectrum CF2O 1100 1150 1200 1250 Wavenumber (cm-1) 1300 1350 Figure 2: Residual IR feature after subtraction of C6F13CH2CH2OH (top) reference spectrum of CF2O 5 residual spectrum C6F13C(O)H Absorbance 1100 1150 1200 1250 Wavenumber (cm-1) 1300 1350 Figure 3: Residual feature after subtraction of C6F13CH2CH2OH (top) reference spectrum of C6F13CHO residual spectrum C6F13CH2OH Absorbance 1120 1160 1200 1240 Wavenumber (cm-1) 1280 Figure 4: Residual feature after subtraction of C6F13CH2CH2OH (top) reference spectrum of C6F13CH2OH 6 0.03 0.02 dilution + wall loss dilution ln([SF6]0/[SF6]) ln([C6F13CH2CH2OH]0/[C6F13CH2CH2OH]) ln([VOC]0/[VOC]) 0.01 28 06 04 EUPHORE C6F13CH2CH2OH + HONO DILUTION AND WALL LOSS 0 0 1000 2000 t (s) 3000 4000 Figure 5: Wall loss of C6F13CH2CH2OH and dilution rate constants: comparison. area (arbitrary unit) C6F13CH2CH2OH (ppbV) HCOOH (ppbV) 4 260 3 240 2 20 C6F13CH2CH2OH HCOOH band 1194-1264 cm-1 16 12 1 220 0 -1 200 -4000 8 4 C6F13CH2CH2OH + HONO 28 06 04 EUPHORE 0 0 4000 8000 time (s) (t = 0: chamber opened) 12000 Figure 6: Concentration - time profiles 7 260 160000 C6F13CH2CH2OH ion 341 unknown 1 x 10 ion 341 unknown 2 240 120000 C6F13CH2CH2OH (ppbV) Unknown products (Area ion 341) 220 80000 200 40000 180 0 0 10000 20000 time (s) t = 0 : chamber opened Figure 7: GC-MS concentration time profiles of C6F13CH2CH2OH and the two unidentified products 1.b. OH-initiated oxidation of C8F17CH2CH2OH (June 29th and 30th) Two runs were performed: in the first one the photolysis of HONO was used as the OH source while H2O2 was used in the second one. The C8F17CH2CH2OH initial concentrations were 130 ppbv and 180 ppbv, respectively. In both experiments the consumption of C8F17CH2CH2OH was estimated to be 20 %. GC-MS analysis showed two mains products, one of them could be the "first" aldehyde C8F17CH2CHO which was supported by HPLC-MS analysis that indicated the formation of a carbonyl compound. Similarly to what was observed for C6F13CH2CH2OH, the FTIR analysis was rapidly complicated by the overlapping IR bands of different fluorinated compounds present in the system (reactant and reaction products). Hence, in absence of IR spectra of pure product samples and reference spectra, it is still difficult to conclude on the identification of the reaction products. 8 0.05 dilution + wall loss dilution 0.04 ln([VOC]0/[VOC]) 0.03 0.02 0.01 29 06 04 EUPHORE C8F17CH2CH2OH + HONO DILUTION AND WALL LOSS 0 0 1000 2000 3000 t (s) 4000 5000 Figure 8: Wall loss of C8F17CH2CH2OH and dilution rate constants: comparison 144 [C8F17CH2CH2OH] 140 ppbV 136 132 C8F17CH2CH2OH + HONO 29 06 04 EUPHORE 128 -4000 0 time (s) (t = 0: chamber opened) 4000 Figure 9: Concentration-time profile of C8F17CH2CH2OH 2. Cl-Initiated oxidation of C6F13CH2OH at EUPHORE (July 1st and 2nd) The aim of these experiments was to generate C6F13CHO and check for its photolysis under sunlight conditions. Chlorine atoms were produced using the photolysis of ClC(O)C(O)Cl. Initial 9 concentrations of C6F13CH2OH were 180 ppbv and 220 ppbv, respectively on July 1st and 2nd. The initial concentrations of ClC(O)C(O)Cl were chosen in order to obtain different consumption of C6F13CH2OH in both experiments (85 % and 15 % in the first and second run, respectively). In both experiments, CF2O was the major product observed by FTIR analysis. Figures 10 show examples of concentration-reaction time profiles for the reactants and products. 2E+014 1.6E+014 1.2E+014 molecule cm-3 8E+013 4E+013 C6F13CH2OH CF2O [CF2O] molecule cm-3 [C6F13CH2OH] molecule cm-3 0 -5000 0 5000 10000 15000 20000 time (s) t = 0 chamber opened 6.2E+012 6E+012 July 1st C6F13CH2OH CF2O 3.0E+011 5.8E+012 5.6E+012 2.0E+011 5.4E+012 5.2E+012 1.0E+011 5E+012 0 4000 time (s) t=0 chamber opened July 2nd 8000 0.0E+000 Figure 10: Concentration-time profiles of C6F13CH2OH and CF2O. 10 The expected mechanism of the Cl-initiated oxidation of C6F13CH2OH leading to the aldehyde C6F13CHO is: Cl + C6F13CH2OH HCl + C6F13CHOH C6F13CHOH + O2 C6F13CHO + HO2 C6F13CHO then reacts with Cl and could also be photolysed: C6F13CHO + Cl C6F13CO + HCl C6F13CHO + h C6F13CO + H C6F13CHO + h C6F13 + HCO C6F13CO and C6F13 radicals will ultimately lead to CF2O. We have used the rate constants values for (k(Cl + C6F13CH2OH) and k(Cl + C6F13CHO)) measured in this work (see next section) to fit the EUPHORE experimental profiles to tentatively assess the importance of the photolysis of the aldehyde. This could not be achieved because the FTIR analysis was rapidly complicated following the formation of the reaction products and the overlapping of the fluorinated compounds with similar IR spectra. The IR bands of the aldehyde could not be clearly distinguished hence the analysis was not clean enough to derive the concentrations of the aldehyde versus reaction time as indicated by the figure 11 below. 11 12 Figure 11: IR spectra obtained during the study of the Cl-initiated oxidation of C6F13CH2OH at EUPHORE using sunlight conditions. 3. Kinetic and mechanism of the Cl reaction with C6F13CH2OH studied in the photoreactor of Wuppertal Experiments were performed in a 480L duran glass reactor interfaced to an in-situ FTIR spectrometer (Nicolet Magna 520) with a pathlength of 51.6 m and a resolution of 1 cm-1. The reactor was surrounded with 20 superactinic lamps (Philips TLA 40W/05, 300 < < 450 nm, max = 360 nm) used to photochimically initiate the experiments. Chlorine atoms were produced by photolysis of molecular chlorine. Relative rate technique was used to measure the rate constant of the reaction of Cl with C6F13CH2OH relative to that of Cl with CH3Cl. The relevant reactions in the system were: Cl + C6F13CH2OH products (1) Cl + CH3Cl products (2) Assuming that the fluoroalcohol and the CH3Cl are consumed only by Cl, it can be shown that: ln([C6F13CH2OH]o/([C6F13CH2OH]t)/ = k1/k2 ln([CH3Cl]o/([CH3Cl]t) where the subscripts 0 and t indicate concentrations before irradiation and time t, respectively. 13 Reactants concentrations were monitored using the IR absorption features 3550-3650 cm-1 and 2820-3150 cm-1 for C6F13CH2OH and CH3Cl, respectively. Infrared spectra were derived from 32 co-added interferograms. Control experiments showed that side reactions such as photolysis of C6F13CH2OH and CH3Cl, as well as heterogenous and dark reactions were negligible in our experimental conditions. Figure 12 shows an example of the obtained data. Using the value of k(Cl+CH3Cl) = 4.8x10-13 cm3 molecule-1 s-1 for the reference reaction, we have derived the reaction rate constant of Cl with C6F13CH2OH: k (Cl + C6F13CH2OH) = (6.5 0.4) x10-13 cm3 molecule-1 s-1 1.2 ln([C6F13CH2OH]0/[C6F13CH2OH]t) 0.8 0.4 0 0 0.2 0.4 0.6 0.8 1 ln([CH3Cl]0/[CH3Cl]t) Figure 12: Loss of C6F13CH2OH versus loss of CH3Cl following the irradiation of the C6F13CH2OH/CH3Cl/air/Cl2 mixture for two independent experiments. The obtained value in this work can be compared with the data reported so far on the reactions of Cl with shorter fluorinated alcohols with the formula CnF2n+1CH2OH as shown (Table 1) which shows that that the length of CnF2n+1 group does not effect on the reactivity of Cl atoms with CnF2n+1CH2OH (k(Cl+CnF2n+1CH2OH) 6.5x10-13 cm3 molecule-1 s-1 for n = 1 up to n = 6). 14 Table 1: Rate constant values for the reaction of Cl with fluoroalcohols Fluoroalcohol CF3CH2OH CF3CF2CH2OH CF3(CF2)2CH2OH CF3(CF2)3CH2OH CF3(CF2)5CH2OH k(Cl) (6.5 0.5) x10-13 (6.5 0.5) x10-13 (6.5 0.5) x10-13 (6.5 0.5) x10-13 (6.5 0.4) x10-13 Product Study of the Cl reaction with C6F13CH2OH: The Cl-initiated oxidation of C6F13CH2OH was investigated using the same system as that used for the kinetic study. The reaction of Cl with C6F13CH2OH proceeds by abstraction of H-atom from -CH2- group followed by reaction with O2 leading to the fluoroaldehyde C6F13CHO, as already mentioned : Cl + C6F13CH2OH C6F13CHOH + HCl C6F13CHOH + O2 C6F13CHO + HO2 The produced aldehyde may undergo further reaction with Cl atoms leading to other fluorinated compounds as end products. Figure 13 shows an example of the obtained experimental concentration profile of C6F13CHO versus C6F13CH2OH. 0.16 22 11 04 0.12 [C6F13C(O)H] t /[C6F13CH2OH] 0 0.08 0.04 0 0 Fit 4: fit aldehyde Equation Y = (A/B)*X*(POW( X, B )-1) B = 3.340333806 A = -0.9801463796 Number of data points used = 60 Residual sum of squares = 0.000276506 Coef of determination, R-squared = 0.997656 Line/Symbol Plot 1 Fit 4: fit aldehyde 0.2 0.4 0.6 0.8 1 [C6F13CH2OH] t /[C6F13CH2OH]0 Figure 13: Formation of C6F13CHO versus loss of C6F13CH2OH 15 The reaction rate constant of Cl with C6F13CHO was derived from the fitting to these profiles: k(Cl + C6F13CHO) = (2.8 0.7)x10-12 cm3 molecule-1 s-1 PUBLICATIONS T. KELLY, V. BOSSOUTROT, I. MAGNERON, K. WIRTZ, J. TREACY, A. MELLOUKI, H. SIDEBOTTOM, G. LE BRAS A kinetic and mechanistic study of the reactions of OH radicals and Cl atoms with 3,3,3trifluoropropanol under atmospheric conditions J. Phys. Chem. A (2005) 7, 334-341 A MELLOUKI, G. SOLIGNAC, G. LE BRAS, I. BARNES, R.L. WATERLAND The atmospheric chemistry of n-C6F13CH2OH To be presented at the SETAC Europe 15th Annual Meeting, Lille, 22-26 May 2005. (This work will also be submitted for publication) 16