Temperature Effects on the Removal of Potential HFC Replacements

Apr 15, 2011 - Camilo José Cela, s/n. 13071 Ciudad Real, Spain. Environ. Sci. Technol. , 2011, 45 (10), pp 4323–4330. DOI: 10.1021/es103931s. Publicat...
0 downloads 0 Views 1MB Size
Download PDF
ARTICLE pubs.acs.org/est

Temperature Effects on the Removal of Potential HFC Replacements, CF3CH2CH2OH and CF3(CH2)2CH2OH, Initiated by OH Radicals María Anti~nolo, Elena Jimenez, and Jose Albaladejo Departamento de Química Física, Facultad de Ciencias Químicas, Universidad de Castilla-La Mancha, Avda. Camilo Jose Cela, s/n. 13071 Ciudad Real, Spain ABSTRACT: The gas-phase kinetic coefficients of OH radicals with two primary fluorinated alcohols, CF3CH2CH2OH (k1) and CF3(CH2)2CH2OH (k2), potential replacements of hydrofluorocarbons (HFCs), are reported here as a function of temperature (T = 263358 K) for the first time. k1 and k2 (together referred as ki) were measured under pseudo-firstorder conditions with respect to the initial OH concentration using the pulsed laser photolysis/laser induced fluorescence technique. The observed temperature dependence of ki (in cm3 molecule1 s1) is described by the following Arrhenius expressions: k1(T) = (2.82 ( 1.28)  1012 exp{(302 ( 139)/T} cm3 molecule1 s1 and k2(T) = (1.20 ( 0.73)  1011 exp{(425 ( 188)/T} cm3 molecule1 s1.The uncertainties in the Arrhenius parameters are at a 95% confidence level ((2σ). Uncertainties in ki(T) include both statistical and systematic errors. Activation energies were (2.5 ( 1.2) kJ/mol and (3.6 ( 1.6) kJ/mol for the OH-reaction with CF3CH2CH2OH and CF3(CH2)2CH2OH, respectively. The global lifetime (τ) at 275 K for CF3CH2CH2OH and CF3(CH2)2CH2OH due to the OH-reaction was estimated to be ca. 2 weeks and 5 days, respectively. The reported Arrhenius parameters can be used in 3D models that take into account the geographical region and season of emissions for estimating a matrix of instantaneous lifetimes. As a consequence of the substitution of the CH3 group by a CH2OH group in HFCs, such as CF3CH2CH3 and CF3(CH2)2CH3, the tropospheric lifetime with respect to the OH reaction is significantly shorter and, since their radiative forcing is similar, global warming potentials of CF3CH2CH2OH and CF3(CH2)2CH2OH are negligible. Therefore, CF3CH2CH2OH and CF3(CH2)2CH2OH seem to be suitable alternatives to HFCs.

1. INTRODUCTION Hydrofluorocarbons (HFCs) were considered to be acceptable alternatives to chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons (HCFCs) on a long-term basis, because they do not directly affect stratospheric ozone. However, the potential contribution of these ozone-friendly substances to global warming is high.1 Recently, hydrofluoroethers (CF3(CF2)xg0O(CH2)yg0CH3, HFEs) and partially fluorinated alcohols (CF3(CH2)xg0CH2OH, HFAs) have been suggested as substitutes for HFCs in a wide range of applications, such as refrigeration, carrier fluids for lubricants, and cleaning of electronic components.2 An assessment of the environmental impact of potential HFC replacements requires information concerning their atmospheric persistence, degradation products, and global warming potentials (GWPs). The main degradation route for HFAs is expected to be their tropospheric removal initiated by hydroxyl (OH) radicals. The lifetimes (and GWPs relative to CO2 for a 20 year horizon time) for CF3CH2OH, CF3CH2CH2OH, and CF3(CH2)2CH2OH have been reported to be 117 days (61.4), 12 days (9.5), and 4 days (1.8), respectively.3,4 The temperature reduction with the altitude in the troposphere, which alters chemical reaction rates, is a significant meteorological parameter in modeling atmospheric pollutant r 2011 American Chemical Society

concentrations. Thus, the knowledge of the T-expression for the OH-rate coefficient allows modeling the pollutant profiles in the troposphere over a wide range of temperatures, corresponding to different seasons, altitudes, and locations. Bearing this in mind, we present here the gas-phase OH-kinetic study for two potential CFC alternatives between 263 and 358 K. OH þ CF3 CH2 CH2 OH f Products k1 ðTÞ

ðR1Þ

OH þ CF3 ðCH2 Þ2 CH2 OH f Products k2 ðTÞ

ðR2Þ

To our knowledge, only room temperature measurements of k1 and k2 have been reported up to now.46 Therefore, this work constitutes the first temperature dependence study on the OHrate coefficient for reactions R1 and R2. The reported rate coefficients k1(T) and k2(T) could be used in 3D models that take into account the geographical region and season of emissions of CF3CH2CH2OH and CF3(CH2)2CH2OH for estimating a matrix of instantaneous lifetimes. The effect of the Received: November 23, 2010 Accepted: April 4, 2011 Revised: March 31, 2011 Published: April 15, 2011 4323

dx.doi.org/10.1021/es103931s | Environ. Sci. Technol. 2011, 45, 4323–4330

Environmental Science & Technology

ARTICLE

Table 1. Experimental Conditions Employed in the Kinetic Study of the Reaction of OH with CF3CH2CH2OH Together with the Pseudo-First Order Rate Coefficient Ranges and the Absolute Rate Coefficient Obtained at Each Temperature [HFA]/ 14

-3

T/ K

pT/ Torr

FT/ sccm

10 cm

263 270

4971 4971

322538 317538

0.75.0 0.74.8

Eλ/ mJ

[H2O2]/ 14

-3

-1

k1(T)  1012 /

[OH]0 / -2

11

-3

10 cm

pulse cm

10 cm

0.60.8 0.71.0

6.97.8 7.07.8

1.11.3 1.21.7

-1

0

-1

k0 / s

k1 /s

cm3 molecule-1 s-1

101127 110166

160530 210540

0.88 ( 0.09 0.87 ( 0.10

278

4971

317538

0.75.1

0.41.4

6.37.8

0.62.5

62236

150590

1.03 ( 0.13

287

4971

316538

0.65.0

0.50.7

6.07.7

0.81.2

87120

190560

1.03 ( 0.11 1.01 ( 0.11

308

4781

266435

0.76.4

2.93.6

5.37.2

4.05.2

501619

6001140

323

4593

265540

0.65.6

2.73.1

5.37.2

3.44.6

474552

6101120

1.13 ( 0.12

338

6080

266435

0.76.9

2.33.3

4.97.3

2.64.7

422589

4801170

1.10 ( 0.12

358

4992

318489

0.75.6

2.02.3

4.97.3

2.33.5

380434

5101050

1.25 ( 0.13

substitution of the CH3 group in HFC molecules by a  CH2OH group on the OH-rate coefficient is discussed. The activation energies found for the title reactions are compared with those for the OH-reactions with other fluorinated alcohols. Lastly, the possible atmospheric implications of the emission of CF3CH2CH2OH and CF3(CH2)2CH2OH are discussed in terms of the tropospheric lifetime and the possible reaction products.

2. EXPERIMENTAL SECTION 2.1. Description of the Experimental System. The setup employed to determine ki has been extensively described hitherto.4,712 Thus, only a brief description is given below. The 248nm radiation from a KrF exciplex laser (Optex Lambda Physik, at a repetition rate of 10 Hz) photodissociates H2O2(g) to form the transient species, OH radicals, at the center of a 200-cm3 Pyrex reaction cell.

H2 O2 þ hνðλ ¼ 248 nmÞ f 2OH

ðR3Þ

The initial OH concentration, [OH]0, was estimated by using the measured laser fluence, E248 nm (= 4.97.8 mJ pulse1 cm2, hence the number of photons cm2 pulse1 at 248 nm, N), the absorption cross section of H2O2 at 248 nm (σλ), the quantum yield for OH production at the photolysis wavelength from H2O2 (φλ),13 and the concentration of the precursor. ½OH0 ¼ Nσ λ φλ ½H2 O2 0

ðE1Þ

Ground state OH(X2Π) radicals were probed to the first excited state (A2Σþ) at 282 nm by a frequency-doubled dye laser, which is pumped by a Nd:YAG laser. The laser-induced A-X (0,0) fluorescence (LIF) of OH radicals was detected by a photomultiplier tube (PMT). A bandpass filter with a maximum transmission at λmax = 350 and 150 nm fwhm was placed in front of the PMT to collect the OH radical LIF emission between 282 and 343 nm. Temporal synchronization of both lasers by means of delay generators allowed varying the reaction time, t, and recording the decay profile of the LIF signal at time scales between 3 and 10 ms.

2.2. Methodology. Gaseous H2O2 (hydrogen peroxide) was introduced into the reactor by bubbling He (helium) through a preconcentrated aqueous solution of H2O2 as described by Jimenez et al.10 Diluted mixtures of HFAs in He were prepared in a 10-L blackened bulb with dilution factors (f = pHFA/{pHFA þ pHe}) ranging from 5  103 to 1.2  102. Partial pressures of HFAs in the bulb ranged from 5.3 to 8.0 Torr for CF3CH2CH2OH and from 3.0 to 3.5 Torr for CF3(CH2)2CH2OH. An additional flow of He was introduced in the reaction cell to get the desired total pressure, pT. All gases were introduced into the reaction cell by means of calibrated mass flow controllers. The mass flowmeters were calibrated at 298 K by introducing the gas (He and HFAs/He gas mixtures) at a fixed flow rate in a bulb of a known volume (3063 mL). The increase of pressure in the bulb was measured to derive the flow rates by means of a capacitance transducer (Leybold Ceravac 0100 Torr). An accurate knowledge of [HFA] and the uncertainties associated with its determination is needed to derive a reliable rate coefficient. As in previous works, an attempt to optically measure [HFA] was made using the UV spectroscopy at 185 nm and Fourier transform infrared (FTIR) spectroscopy. Both experimental set-ups have been previously described.4,7,10 The measured absorption cross sections (in base e) at 185 nm, σλ=185 nm, of CF3CH2CH2OH and CF3(CH2)2CH2OH were 5.16  1019 and 4.96  1019 cm2 molecule1, respectively. No absorption from the HFAs was detected at 185 nm over the concentration range employed. In the case of the FTIR spectroscopy, the integrated absorption cross sections, Sint, for the OH stretching band of CF3CH2CH2OH and CF3(CH2)2CH2OH were measured to be 2.25  1018 (3759 3584 cm1) and 1.05  1018 cm molecule1 (37253628 cm1), respectively. Even though using the maximum path length (800 cm) allowed by the multipass FTIR cell, the absorption by CF3CH2CH2OH and CF3(CH2)2CH2OH was too low to accurately measure [HFA] at the levels employed in this kinetic study. For that reason, the HFA concentration in the reaction cell was calculated from the total flow rate (FT), the flow rate of the diluted mixture of HFA (FHFA), the temperature, and pT.

½HFAT ¼ 3:24  1016 molecule cm3 Torr1 pT ðTorrÞf FT (263540 sccm  standard cm3 per minute) in the reaction cell is the sum of all rate flows, i.e., FHFA (110 sccm), FHe through the H2O2 bubbler (210 sccm), and main He

FHFA ðsccmÞ 298 K FT ðsccmÞ TðKÞ

ðE2Þ

(260520 sccm). The [HFA] range was limited by the vapor pressure of these compounds and the total pressure in the dilution bulb (above 500 Torr) needed to keep the mass flow 4324

dx.doi.org/10.1021/es103931s |Environ. Sci. Technol. 2011, 45, 4323–4330

Environmental Science & Technology

ARTICLE

Table 2. Experimental Conditions Employed in the Kinetic Study of the Reaction of OH with CF3(CH2)2CH2OH Together with the Pseudo-First Order Rate Coefficient Ranges and the Absolute Rate Coefficient Obtained at Each Temperature [HFA]/ 14

T/ K

pT/ Torr

FT/ sccm

10 cm

263 270

6079 5171

371484 263536

0.63.4 0.43.6

Eλ/ mJ

[H2O2]/

-3

14

-3

-1

k2(T)  1012 /

[OH]0 / -2

11

-3

10 cm

pulse cm

10 cm

0.91.3 0.81.2

6.66.9 5.06.6

1.32.1 1.11.3

0

-1

-1

k0 / s

k2 /s

cm3 molecule-1 s-1

145212 125187

2801050 2701120

2.41 ( 0.25 2.63 ( 0.27

278

5171

263432

0.53.6

0.81.0

6.27.0

1.21.4

123165

2401030

2.48 ( 0.30

287

5282

267384

0.64.0

0.71.3

5.06.5

0.81.8

114211

3001100

2.42 ( 0.26 3.14 ( 0.32

308

4695

265488

0.43.1

2.53.0

5.87.1

3.74.7

438519

6201450

323

6294

317488

0.43.0

2.02.7

5.87.0

2.83.9

357485

6101400

3.53 ( 0.37

338

5475

320437

0.42.6

2.53.3

6.07.1

3.84.5

445601

5901390

3.34 ( 0.34

358

5181

265436

0.32.5

1.72.4

6.07.1

2.73.8

325444

5401310

3.54 ( 0.37

controllers operational. The experimental conditions employed are summarized in Tables 1 and 2 for the OH þ CF3CH2CH2OH and the OH þ CF3(CH2)2CH2OH reactions, respectively. 2.3. Kinetic Data Analysis. The OH LIF signal decays due to the reaction with the HFA (R1 or R2), the reaction with the photochemical precursor (R4), and the diffusion out of the detection zone (R5). OH þ H2 O2 f H2 O þ HO2

k4

OH f Loss k5

ðR4Þ ðR5Þ

Under pseudo-first order conditions ([HFA]0 and [H2O2]0 . [OH]0), the decay of the OH LIF signal, ILIF, is described by a single exponential in the absence of any secondary chemistry. ILIF ðtÞ ¼ ILIF ðt ¼ 0Þexpð  ki 0 tÞ

ðE3Þ

The pseudo-first order decay rate coefficient, ki0 , can be obtained from the nonlinear least-squares analysis of the OH LIF signal. ki0 is given by the following expression: ki 0 ¼ ki ½HFA þ k4 ½H2 O2  þ k5 ¼ ki ½HFA þ k0

ðE4Þ

where k0 is the rate coefficient determined in the absence of HFA. Equation E4 can be also written as: ki 0  k0 ¼ ki ½HFA

ðE5Þ

Thus, the bimolecular rate coefficients ki can be obtained from the slope of plots of the eqs E4 or E5 (see Section 3). As described hitherto,12 upper limits of the H2O2 concentration were derived by neglecting k5 in k0 ([H2O2]0 = (0.43.6)  1014 molecule cm3). By using eq E1, the upper limit of [OH]0 ranged from 6  1010 cm3 to 5.2  1011 cm3. Typically, [OH]0 ranged from 1  1011 cm3 to 4  1011 cm3. A numerical simulation using Facsimile program was carried out to check the possible influence of the OH þ OH and OH þ HO2 reactions on ki0 . At the OH initial concentrations used, radicalradical reactions did not significantly contribute to the OH decay. From a previous analysis performed by headspace chromatography coupled to mass spectrometry detection,4 the impurities present in the HFA samples were ethanol (0.4% in CF3CH2CH2OH) and CF3CHdCH2 (3% in CF3(CH2)2CH2OH). Reagents. He (99.999%, Praxair) was used as supplied. Purities of the liquid samples were 98% for CF3CH2CH2OH (Apollo Scientific Ltd.), and 97% for CF3(CH2)2CH2OH (Apollo Scientific Ltd.) and H2O2 (Sharlab, 50% w/v). H2O2 was preconcentrated as described by Jimenez et al.10 Organic liquid samples were degassed by several freezepumpthaw cycles.

Figure 1. Decay temporal profiles of the LIF OH signal for CF3CH2CH2OH (9, 1.5  1014 cm3) and CF3(CH2)2CH2OH (O, 1.6  1014 cm3) at 263 K (a) and 358 K (b).

3. RESULTS AND DISCUSSION 3.1. Determination of ki. Some examples of the temporal profiles (in log scale) of ILIF are shown in Figure 1. No curvature in the ln(ILIF) versus t plots was observed in the 10-ms time scale, 4325

dx.doi.org/10.1021/es103931s |Environ. Sci. Technol. 2011, 45, 4323–4330

Environmental Science & Technology

Figure 2. Pseudo-first-order decay rate versus the concentration of CF3CH2CH2OH (empty symbols, x = 1) and CF3(CH2)2CH2OH (full symbols, x = 2) at low and high temperatures.

being linear over 25 OH-lifetimes. ki0 and k0 extracted from the analysis of the ILIF decay according to eq E3 are listed in Tables 1 and 2 for CF3CH2CH2OH and CF3(CH2)2CH2OH, respectively. Any uncertainty in [HFA] and/or ki0 directly affects the reported ki. To check the possible systematic errors associated with the contribution of the OH þ H2O2 reaction to ki0 , the kinetic analysis was first performed according to eq E4, without including the experimental k0. The intercepts of ki0 versus [HFA] plots were consistent with the measured k0. Therefore, all data recorded at a given temperature were combined and analyzed according to eq E5. Even though the contribution of k0 to ki0 was larger than 10% in all cases, the rate coefficients ki(T) obtained from both kinetic analyses are in excellent agreement, indicating that no systematic uncertainties seem to be associated with the contribution of the OH þ H2O2 reaction. Tables 1 and 2 summarize the obtained k1 and k2 as a function of temperature. In addition, if a very reactive impurity is present in a significant concentration, ki0 could be overestimated. At the impurity levels cited, the influence on k10 and k20 was found to be negligible (less than 2%) under the experimental conditions used. A possible OH formation from CF3(CH2)xCH2OH photolysis was also investigated in experiments carried out in the absence of H2O2. OH radicals were not detected, indicating that photolysis of HFA at 248 nm is not expected to contribute to any OH regeneration or to a significant HFA loss at the laser fluences employed. At high temperatures, the OH regeneration by the thermal decomposition of CF3(CH2)xCHOH radicals formed in reactions R1 and R2 could yield an underestimation of ki0 , and therefore in ki. However, this possibility was ruled out by Rajakumar et al.14 by studying OD/OH exchange in the reactions with CFH2CH2OH at 355 K. 3.2. Temperature Dependence of Rate Coefficients ki. The rate coefficients for reactions R1 and R2 at each temperature were obtained by plotting all data recorded at different total pressures (in the 4595 Torr range). Figure 2 shows several examples of the plot of ki0 -k0 versus the HFA concentration at high and low temperatures. Tables 1 and 2 list the obtained k1 and k2 between 263 and 358 K. The estimated systematic

ARTICLE

Figure 3. Arrhenius plot for the OH-reactions of CF3CH2CH2OH and CF3(CH2)2CH2OH. Dashed lines are the bands of uncertainty according to the JPL/NASA data evaluation format.

uncertainties in the HFA concentrations in the reaction cell were derived by propagating the errors in total pressure, temperature, and flow rates ((3%). Considering the uncertainties associated with the measurements of [HFA] and ki0 , a systematic error of (4% in k1 and (5% in k2 was estimated. Conservatively, (10% ((2σ) was quadratically added to the statistical errors obtained from the fit to eq E5. The analysis of the experimental data by nonlinear leastsquares fitting to the Arrhenius equation results in the following expressions for the observed T-dependence of k1 and k2: k1 ðT ¼ 263  358 KÞ ¼ ð2:82 ( 1:28Þ  1012 expf  ð302 ( 139Þ=Tg cm3 molecule1 s1

ðE6Þ k2 ðT ¼ 263  358 KÞ ¼ ð1:20 ( 0:73Þ  1011 expf  ð425 ( 188Þ=Tg cm3 molecule1 s1

ðE7Þ Uncertainties in the Arrhenius parameters (A and Ea/R) are at 95% confidence level ((2σ). The uncertainty at each temperature can also be expressed with the format of the JPL/NASA data evaluation:15   1 1 ðE8Þ  f ðTÞ ¼ f ð298 KÞexp g T 298 The estimated uncertainties are f1 (298 K) = 1.12 and g1 = 100 K for CF3CH2CH2OH and f2 (298 K) = 1.10 and g2 = 90 K for CF3(CH2)2CH2OH. In Figure 3 k1(T) and k2(T) are displayed together with the Arrhenius expressions (solid lines) and the band of uncertainties (dashed lines). Room temperature values of ki reported by Jimenez et al.4 (k1 (298 K) = 9.70  1013 cm3 molecule1 s1 and k2 (298 K) = 2.62  1012 cm3 molecule1 s1) are also depicted in Figure 3 and were included in the fitting procedure. For reactions R1 and R2, positive 4326

dx.doi.org/10.1021/es103931s |Environ. Sci. Technol. 2011, 45, 4323–4330

Environmental Science & Technology

ARTICLE

Table 3. Comparison of the Temperature Dependence of kOH for Several C2C4 Fluorinated and Non-Fluorinated Alcohols kOH (298 K)  1012/cm3 molecule-1 s-1

A  1012/cm3 molecule-1 s-1

(Ea/R)/K

CH3CH2OH

3.2

3.0

20

evaluation b

CF3CH2OH

0.1

1.9

863

evaluation c

alcohol

technique a

C2

0.107 ( 0.003

d

PLPRF

0.123 ( 0.006 e

RRFTIR 2.00 ( 0.37 f

0.107 ( 0.005 f

CF2HCH2OH

890 ( 0.60 f

PLPLIF

0.0968 ( 0.0023 f

FPLIF

0.095 ( 0.0071 f 0.252 ( 0.044 d

FPRF PLPRF

0.451 ( 0.004 e

RRFTIR SAR estimation f

0.11 CFH2CH2OH

DFLIF

0.098 ( 0.0041 f

1.42 ( 0.11 e

RRFTIR 5.15 ( 0.88 g

1.63 ( 0.09 g

330 ( 45 g

PLPLIF SAR estimation f

0.29 C3 CH3CH2CH2OH CF3CH2CH2OH

5.8 0.97 ( 0.11 h

4.6 2.84 ( 1.21 i

70 306 ( 130 i

0.691 ( 0.091 j

evaluation b PLPLIF RR/FTIR

1.08 ( 0.05 k

RR/GCFID/FTIR

0.89 ( 0.03 k

PLPLIF

2.47

SAR estimation f

CF2HCH2CH2OH

2.49

SAR estimation f

CFH2CH2CH2OH

2.52

CF3CF2CH2OH

SAR estimation f 2.27þ1.15l 0.76

l

0.111 0.102 ( 0.04 f

900 ( 70 780 ( 60 f l

1.40 ( 0.27 f

RRGC/MS DF/PLPLIF

0.102 ( 0.01 m

RR/FTIR

CH2FCF2CH2OH

0.107

SAR estimation

CHF2CF2CH2OH

0.106

SAR estimation

C4 CH3(CH2)2CH2OH CF3(CH2)2CH2OH CHF2(CH2)2CH2OH CH2F(CH2)2CH2OH

8.5 2.62 ( 0.63

5.3 12.2 ( 0.73

h

evaluation b

- 140 i

434 ( 184

i

PLPLIF

2.91

SAR estimation

2.93 2.99

SAR estimation SAR estimation

a FP, flash photolysis; PLP, pulsed-laser photolysis; LIF, laser induced fluorescence; RF, resonance fluorescence; RR, relative rate; FTIR, Fourier transform infrared detection; GC/MS, gas chromatography/mass spectrometry; DF, discharge flow tube; FID, flame ionization detection. b IUPAC  evaluation.28 c IUPAC evaluation.17 d Kovacs et al.18 e Sellevag et al.3 f Tokuhashi et al.21 T = 250430 K . g Rajakumar et al.14 T = 238355 K. h Jimenez 4i j 5k 6l et al. This work. Hurley et al. Kelly et al. Chen et al.22 T = 298356 K. m Hurley et al.29

activation energies, Ea ((2σ) in kJ/mol, were determined to be (2.5 ( 1.2) and (3.6 ( 1.6), respectively. Table 3 compares the OH-rate coefficient at room temperature (kOH (298 K)) and A and Ea/R for several C2C4 fluorinated and nonfluorinated alcohols. In primary alcohols, the substitution of the CH3 group by a CF3 group has a great impact in both kOH and Ea. The CF3 group strongly deactivates the H-atom abstraction by OH radicals in HFAs (see Table 3). For example, kOH (298 K) for CF3CH2OH is ca. 30 times lower than that for CH3CH2OH.16 The observed decrease in kOH (298 K) for CF3CH2CH2OH and CF3(CH2)2CH2OH with respect to that for CH3CH2CH2OH and CH3(CH2)2CH2OH, respectively, is less pronounced than that for CF3CH2OH with respect to ethanol.

The substitution of fluorine atoms by hydrogen atoms in the CF3 group of C2HFAs also significantly affects the reactivity of HFAs toward OH. Thus, kOH (298 K) for CF3CH2OH is 1 order of magnitude lower than that for CFH2CH2OH.3,14,17,18 To our knowledge, no kinetic data on the reactions of OH radicals with CF2HCH2CH2OH and CFH2CH2CH2OH are available to discuss the effect of F-substitution at the γ-carbon atom on the OH reactivity of C3HFAs. The structureactivity relationship (SAR) method developed by Kwok and Atkinson19 and modified by DeMore20 and Tokuhashi et al.21 can be employed to calculate kOH (298 K) for those C3HFAs. However, as can be seen in Table 3, SAR estimations of kOH (298 K) are in poor agreement with the experimental rate coefficients. Moreover, SAR method does not differentiate 4327

dx.doi.org/10.1021/es103931s |Environ. Sci. Technol. 2011, 45, 4323–4330

Environmental Science & Technology

ARTICLE

among the reactivities of CF3CH2CH2OH, CF2HCH2CH2OH, and CFH2CH2CH2OH or between the reactivities of CFH2CF2CH2OH and CF2HCF2CH2OH. The substitution of the CH2 group of the β-carbon in CF3CH2CH2OH by a CF2 group reduces the OHreactivity in almost 1 order of magnitude.21,22 The effect of adding methylene groups to the fluorinated alcohol molecule is also shown in Table 3. The OH-rate coefficient increases from 1  1013 cm3 molecule1 s1 for CF3CH2OH17 to 2.62  1013 cm3 molecule1 s1 for CF3(CH2)2CH2OH,4 indicating that multiple H-atom abstraction reaction pathways are favored. H-atom abstraction can occur from the hydrogen of the CH2 group adjacent to the OH group (channel R6a), from the alcohol group (channel R6b), or from the hydrocarbon chain (channel R6c). OH þ CF3 ðCH2 Þx CH2 OH f H2 O þ CF3 ðCH2 Þx CHOH ðR6aÞ

f H2 O þ CF3 ðCH2 Þx CH2 O f H2 O þ CF3 ðCH2 Þx  1 CHCH2 OH

ðR6bÞ ðR6cÞ

In the presence of O2, the main degradation product of CF3CH2CH2OH initiated by Cl atoms has been reported to be CF3CH2CHO by Papadimitriou et al.23 and Hurley et al.5 These observations indicate that the H-atom abstraction by Cl atoms from the CH2 group attached to the OH group is favored. A similar behavior could be expected for the OH reactions. However, as OH radicals are more selective than Cl atoms, channel R6c may not be negligible. As shown in Table 3, the activation energy for the OHreactions with ethanol, 1-propanol, and 1-butanol was reported to be slightly negative. In contrast, Ea becomes positive for the OH-reactions with the corresponding HFAs and perfluoroalcohols (PFAs), suggesting that these reactions involve a direct H-abstraction by OH radicals. The reported activation energies for HFAs and PFAs are slightly positive and range from 2.5 to 7.5 kJ per mole. Wang et al.24 calculated the rate coefficients for the reaction of OH with CF3CF2CH2OH between 200 and 2000 K by canonical variational transition state theory. These authors reported a curvature in the Arrhenius plot as consequence of the presence of two reaction channels. A deviation of the Arrhenius behavior is expected for reactions R1 and R2 if multiple H-atom abstraction reaction pathways exist. Unfortunately, the limited temperature range employed did not allow the observation of any curvature in the Arrhenius plots, even for the OH-reaction with CF3(CH2)2CH2OH. 3.3. Atmospheric Fate of CF3(CH2)xCH2OH. As a first approximation, lifetimes for CF3CH2CH2OH and CF3(CH2)2CH2OH, τ, can be calculated by considering an average tropospheric temperature and OH concentration of 1  106 cm3.25 τ¼

1 ki ð275 KÞ½OHavg

ðE9Þ

where ki (275 K) is the OH-rate coefficient for the title HFAs at that temperature. For comparative purposes, this simplified calculation of lifetime is commonly used and reported to be adequate for shortlived atmospheric compounds.1 Taking into account the T-dependence of k1 and k2 given by eqs E6 and E7, lifetimes were calculated to be ca. 2 weeks for CF3CH2CH2OH and 4.5 days for CF3(CH2)2CH2OH. Lifetimes calculated from eq E9 provide only a crude estimation of the global mean lifetimes, because of the

significant regional variations in the OH radical concentration, solar flux, and the spatial and seasonal distributions of their sources. Nevertheless, the reported Arrhenius parameters can be used in 3D models that take into account the geographical region and season of emissions for estimating a matrix of instantaneous lifetimes. Photolysis in the actinic region (λ > 290 nm) and removal by reaction with Cl atoms of CF3CH2CH2OH and CF3(CH2)2CH2OH were reported to be negligible.4 Similar to other fluorinated alcohols, direct airwater partition is expected to be an unimportant tropospheric sink for CF3CH2CH2OH and CF3(CH2)2CH2OH. Chen et al.26 reported that molar ratios of CF3CH2OH, CHF2CF2CH2OH, and CF3CF2CH2OH in cloudwater and rain were less than 2.1% in the atmosphere. Consequently, the removal of these alcohols by wet deposition was insignificant. As far as we know, Henry’s law constants, kH,cp, for CF3CH2CH2OH and CF3(CH2)2CH2OH have not been reported. Assuming that the temperature dependence of kH,cp for these HFAs is similar to that reported for CHF2CF2CH2OH by Chen et al.,26 the calculated lifetime due to the wet deposition ranges from 4 months at 5.5 km (altitude in the troposphere where the temperature is close to the lower limit of the experimental temperature range, 260 K) and 10 months at 2 km (close to the planetary boundary layer). The average precipitation rate used in the calculation was 300 mm/yr as described in ref 8. The averaged contribution of this degradation route to the total tropospheric loss of CF3CH2CH2OH and CF3(CH2)2CH2OH would be less than 5%. Thus, this degradation route can also be neglected unless their kH,cp was much higher than assumed and the annual precipitation rates in the emission source was higher than 300 mm/yr. Consequently, the main sink for these HFAs is the homogeneous reaction with OH radicals. The lifetime of the corresponding HFC for CF3CH2CH2OH, CF3CH2CH3 (HFC263fb), is reported to be 1.2 years by the 2010 WMO/UNEP Scientific Assessment of Ozone Depletion.1 For CF3(CH2)2CH3, the corresponding HFC for CF3(CH2)2CH2OH, no kinetic data were found in the bibliography, but according to the structureactivity relationship (SAR) method proposed by Tokuhashi et al.21 the predicted OH-rate coefficient is 7.72  1013 cm3 molecule1 s1. For this HFC, the estimated lifetime at the surface (T = 298.15 K) was calculated to be ca. 15 days. The radiative forcing (RF) of HFC-263fb was reported by Rajakumar et al.27 to be 0.13 W m2 ppbv1. For CF3CH2CH2OH and CF3(CH2)2CH2OH, RF = 0.20 and 0.11 W m2 ppbv1, respectively.4 Although RF for CF3(CH2)2CH3 has not been reported, as far as we know, its value should be similar to those for CF3CH2CH3, CF3CH2CH2OH, and CF3(CH2)2CH2OH. Therefore, as a consequence of the substitution of a CH3 group in CF3CH2CH3 and CF3(CH2)2CH3 by a CH2OH group, the tropospheric lifetime with respect to the OH reaction is significantly shorter and, thus, their direct contribution to the global warming should be negligible.

4. IMPLICATIONS The present study constitutes the first determination of the temperature dependence of the overall rate coefficient for the (k2). These hydrofluoroalcohols are rapidly removed from the troposphere by reaction with OH radicals. Once emitted to the atmosphere their residence times are estimated to be approximately two weeks for CF3CH2CH2OH and several days for CF3(CH2)2CH2OH. This time scale is not long enough to let 4328

dx.doi.org/10.1021/es103931s |Environ. Sci. Technol. 2011, 45, 4323–4330

Environmental Science & Technology these HFAs be homogeneously distributed in the troposphere, but it is also too short compared with the vertical mixing ratio (several months) to let these HFAs reach the stratosphere. Therefore, these HFAs would be degraded close to the emission source or in the nearest area. The relatively short lifetimes of CF3CH2CH2OH and CF3(CH2)2CH2OH make their potential contribution to the global warming of Earth negligible. Thus, CF3CH2CH2OH and CF3(CH2)2CH2OH seem to be suitable alternatives to HFCs.

’ AUTHOR INFORMATION Corresponding Author

Phone: 34 926 29 53 27; fax: 34 926 29 53 18; e-mail: [email protected].

’ ACKNOWLEDGMENT The research described in this article has been funded by the Spanish Ministerio de Ciencia e Innovacion and the Junta de Comunidades de Castilla-La Mancha (CGL2010-19066-61835/ CLI and PCI08-0123-0381 projects, respectively). M.A. thanks the first institution for providing her a fellowship. ’ REFERENCES (1) WMO/UNEP. 2010: Scientific Assessment of Ozone Depletion; 2010. (2) U.S. Environmental Protection Agency (EPA). Clean Air Act. http://www.epa.gov/ozone/snap/index.html.  (3) Sellevag, S. R.; Nielsen, C. J.; Sovde, O. A.; Myhre, G.; Sundet, J. K.; Stordal, F.; Isaksen, I. S. Atmospheric gas-phase degradation and global warming potentials of 2-fluoroethanol, 2,2-difluoroethanol, and 2,2,2-trifluoroethanol. Atmos. Environ. 2004, 38, 6725–6735. (4) Jimenez, E.; Anti~nolo, M.; Ballesteros, B.; Martínez, E.; Albaladejo, J. Atmospheric lifetimes and global warming potentials of CF3CH2CH2OH and CF3(CH2)2CH2OH. Chem. Phys. Chem. 2010, 11, 4079–4087. (5) Hurley, M. D.; Misner, J. A.; Ball, J. C.; Wallington, T. J.; Ellis, D. A.; Martin, J. W.; Mabury, S. A.; Sulbaek Andersen, M. P. Atmospheric chemistry of CF3CH2CH2OH: Kinetics, mechanisms and products of Cl atom and OH radical initiated oxidation in the presence and absence of NOx. J. Phys Chem. A 2005, 109, 9816–9826. (6) Kelly, T.; Bossoutrot, V.; Magneron, I.; Wirtz, K.; Treacy, J.; Mellouki, A.; Sidebottom, H.; Le Bras, G. 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, 109, 347–355. (7) Anti~nolo, M.; Jimenez, E.; Notario, A.; Martínez, E.; Albaladejo, J. Tropospheric photooxidation of CF3CH2CHO and CF3(CH2)2CHO initiated by Cl atoms and OH radicals. Atmos. Chem. Phys. 2010, 10, 1911–1922. (8) Jimenez, E.; Lanza, B.; Anti~nolo, M.; Albaladejo, J. Photooxidation of leaf-wound oxygenated compounds, 1-penten-3-ol, (Z)-3-hexen-1-ol, and 1-penten-3-one, initiated by OH radicals and sunlight. Environ. Sci. Technol. 2009, 43, 1831–1837. (9) Jimenez, E.; Lanza, B.; Martínez, E.; Albaladejo, J. Daytime tropospheric loss of hexanal and trans-2-hexenal: OH kinetics and UV photolysis. Atmos. Chem. Phys. 2007, 7, 1565–1574. (10) Jimenez, E.; Lanza, B.; Garzon, A.; Ballesteros, B.; Albaladejo, J. Atmospheric degradation of 2-butanol, 2-methyl-2-butanol, and 2,3dimethyl-2-butanol: OH kinetics and UV absorption cross sections. J. Phys. Chem. A 2005, 109, 10903–10909. (11) Jimenez, E.; Ballesteros, B.; Martínez, E.; Albaladejo, J. Tropospheric Reaction of OH with selected linear ketones: Kinetic studies between 228 and 405 K. Environ. Sci. Technol. 2005, 39, 814–820.

ARTICLE

(12) Albaladejo, J.; Ballesteros, B.; Jimenez, E.; Martin, P.; Martinez, E. A PLP-LIF kinetic study of the atmospheric reactivity of a series of C4C7 saturated and unsaturated aliphatic aldehydes with OH. Atmos. Environ. 2002, 36, 3231–3239. (13) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Crowley, J. N.; Hampson, R. F.; Hynes, R. G.; Jenkin, M. E.; Rossi, M. J.; Troe, J.; Wallington, T. J. Evaluated kinetic and photochemical data for atmospheric chemistry: Volume I - gas phase reactions of Ox, HOx, NOx and SOx species. Atmos. Chem. Phys. 2004, 4, 1461–1738. (14) Rajakumar, B.; Burkholder, J. B.; Portmann, R. W.; Ravishankara, A. R. Rate coefficients for the OH þ CFH2CH2OH reaction between 238 and 355 K. Phys. Chem. Chem. Phys. 2005, 7, 2498–2505. (15) Sander, S. P.; Friedl, R. R.; Ravishankara, A. R.; Golden, D. M.; Kolb, C. E.; Kurylo, M. J.; Molina, M. J.; Moortgat, G. K.; Keller-Rudek, H.; Finlayson-Pitts, B. J.; Wine, P. H.; Huie, R. E.; Orkin, V. L. Chemical Kinetics and Photochemical Data for Use in Atmospheric Studies, Evaluation Number 15; JPL Publication 06-2; Jet Propulsion Laboratory, California Institute of Technology: Pasadena, CA, 2006. (16) Jimenez, E.; Gilles, M. K.; Ravishankara, A. R. Kinetics of the reactions of hydroxyl radical with CH3OH and C2H5OH between 235 and 360 K. J. Photochem. Photobiol. A: Chem. 2003, 157, 237–245. (17) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Crowley, J. N.; Hampson, R. F.; Hynes, R. G.; Jenkin, M. E.; Rossi, M. J.; Troe, J.; Wallington, T. J. Evaluated kinetic and photochemical data for atmospheric chemistry: Volume IV  gas phase reactions of organic halogen species. Atmos. Chem. Phys. 2008, 8, 4141–4496. (18) Kovacs, G.; Szasz-Vadasz, T.; Papadimitriou, V. C.; Dobe, S.; Berces, T.; Marta, F. Absolute rate constants for the reactions of OH radicals with CH3CH2OH, CF2HCH2OH and CF3CH2OH. React. Kinet. Catal. Lett. 2006, 87, 129–138. (19) Kwok, E. S. C.; Atkinson, R. Estimation of hydroxyl radical reaction rate constants for gas-phase organic compounds using a structure-reactivity relationship. An update. Atmos. Environ. 1995, 29, 1685–1695. (20) DeMore, W. B. Experimental and estimated rate constants for the reactions of hydroxyl radicals with several halocarbons. J. Phys. Chem. 1996, 100, 5813–5820. (21) Tokuhashi, K.; Nagai, H.; Takahashi, A.; Kaise, M.; Kondo, S.; Sekiya, A.; Takahashi, M.; Gotho, Y.; Suga, A. Measurement of the OH reaction rate constants for CF3CH2OH, CF3CF2CH2OH, and CF3CH(OH)CF3. J. Phys. Chem. A 1999, 103, 2664–2672. (22) Chen, L.; Fukuda, K.; Takenaka, N.; Bandow, H.; Maeda, Y. Kinetics of the gas-phase reaction of CF3CF2CH2OH with OH radicals and its atmospheric lifetime. Int. J. Chem. Kinet. 2000, 32, 73–78. (23) Papadimitriou, V. C.; Papanastasiou, D. K.; Stefanopoulus, V. G.; Zaras, A. M.; Lazarou, Y. G.; Papagiannakopoulos, P. Kinetic study of the reactions of Cl atoms with CF3CH2CH2OH, CF3CF2CH2OH, CHF2CF2CH2OH, and CF3CHFCF2CH2OH. J. Phys. Chem. A 2007, 111, 11608–11617. (24) Wang, Y.; Liu, J.; Li, Z.; Wang, L.; Wu, J.; Sun, C. Direct dynamics study on hydrogen abstraction reaction of CF3CF2CH2OH with OH radical. J. Phys. Chem. A 2006, 110, 5853–5859. (25) Krol, M.; van Leeuwen, P. J.; Lelieveld, J. Global OH trend inferred from methylchloroform measurements. J. Geophys. Res. Atmos. 1998, 103, 10697–10711. (26) Chen, L.; Takenaka, N.; Bandow, H.; Maeda, Y. Henry’s law constants for C2-C3 fluorinated alcohols and their wet deposition in the atmosphere. Atmos. Environ. 2003, 37, 4817–4822. (27) Rajakumar, B.; Portmann, R. W.; Burkholder, J. B.; Ravishankara, A. R. Rate coefficients for the reactions of OH with CF3CH2CH3 (HFC-263fb), CF3CHFCH2F (HFC-245eb), and CHF2CHFCHF2 (HFC-245ea) between 238 and 375 K. J. Phys. Chem. A 2006, 110, 6724–31. (28) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Crowley, J. N.; Hampson, R. F.; Hynes, R. G.; Jenkin, M. E.; Rossi, M. J.; Troe, J.; Wallington, T. J. Evaluated kinetic and photochemical data for atmospheric chemistry: Volume II  gas phase reactions of organic species. Atmos. Chem. Phys. 2006, 6, 3625–4055. 4329

dx.doi.org/10.1021/es103931s |Environ. Sci. Technol. 2011, 45, 4323–4330

Environmental Science & Technology

ARTICLE

(29) Hurley, M. D.; Wallington, T. J.; Sulbaek Andersen, M. P.; Ellis, D. A.; Martin, J. W.; Mabury, S. A. Atmospheric chemistry of fluorinated alcohols: Reaction with Cl atoms and OH radicals and atmospheric lifetimes. J. Phys. Chem. A 2004, 108, 1973–1979.

4330

dx.doi.org/10.1021/es103931s |Environ. Sci. Technol. 2011, 45, 4323–4330