CH3Cl, CH2Cl2, CHCl3, and CCl4: Infrared spectra, radiative efficiencies, and global warming potentials

Journal of Quantitative Spectroscopy & Radiative Transfer 174 (2016) 56–64

Contents lists available at ScienceDirect

Journal of Quantitative Spectroscopy & Radiative Transfer journal homepage: www.elsevier.com/locate/jqsrt

CH3Cl, CH2Cl2, CHCl3, and CCl4: Infrared spectra, radiative efficiencies, and global warming potentials Timothy J. Wallington a,n, Bruno Pasquini Pivesso a, Alane Moura Lira a, James E. Anderson a, Claus Jørgen Nielsen b, Niels Højmark Andersen b, Øivind Hodnebrog c a

Research and Advanced Engineering, Ford Motor Company, Mail Drop RIC-3182, Dearborn, MI 48121-2053, USA CTCC, Department of Chemistry, P.O. Box 1033 Blindern, N-0315 Oslo, Norway c Center for International Climate and Environmental Research-Oslo (CICERO), P.O. Box 1129 Blindern, N-0318 Oslo, Norway b

a r t i c l e in f o

abstract

Article history: Received 24 July 2015 Received in revised form 23 November 2015 Accepted 12 January 2016 Available online 28 January 2016

Infrared spectra for the title compounds were measured experimentally in 700 Torr of air at 295 K and systematically modeled in B3LYP, M06-2X and MP2 calculations employing various basis sets. Calibrated infrared spectra over the wavenumber range 600– 3500 cm
1 are reported and combined with literature data to provide spectra for use in experimental studies and radiative transfer calculations. Integrated absorption cross sections are (units of cm
1 molecule
1): CH3Cl, 660–780 cm
1, (3.89 7 0.19)  10
18; CH2Cl2, 650–800 cm
1, (2.16 7 0.11)  10
17; CHCl3, 720–810 cm
1, (4.08 7 0.20)  10
17; and CCl4, 730–825 cm
1, (6.30 7 0.31)  10
17. CH3Cl, CH2Cl2, CHCl3, and CCl4 have radiative efficiencies of 0.004, 0.028, 0.070, and 0.174 W m
2 ppb
1 and global warming potentials (100 year horizon) of 5, 8, 15, and 1775, respectively. Quantum chemistry calculations generally predict larger band intensities than the experimental values. The best agreement with experiments is obtained in MP2(Full) calculations employing basis sets of at least triple-zeta quality augmented by diffuse functions. The B3LYP functional is found ill-suited for calculating vibrational frequencies and infrared intensities of halocarbons. & 2016 Published by Elsevier Ltd.

Keywords: Methyl chloride Methylene chloride Trichloromethane Tetrachloromethane Radiative efficiency Global warming potential

1. Introduction Chloroalkanes are of interest in atmospheric chemistry because of their ability to contribute to chlorine-catalyzed stratospheric ozone loss and to radiative forcing of climate change. Quantitative infrared absorption spectra are required to evaluate the contribution of chloroalkanes to radiative forcing of climate change. In our recent review of the radiative efficiencies and global warming potentials of halocarbons we noted inconsistencies in the literature n

Corresponding author. E-mail address: [email protected] (T.J. Wallington).

http://dx.doi.org/10.1016/j.jqsrt.2016.01.029 0022-4073/& 2016 Published by Elsevier Ltd.

database for quantitative infrared absorption spectra of the chloromethanes [1]. For example there are approximately 20–30% discrepancies in the intensities of the published spectra for CHCl3 and CCl4. To improve the database of IR spectra of halocarbons for assessments of radiative forcing of climate change we present new results of an experimental and computational study of the IR spectra of CH3Cl, CH2Cl2, CHCl3, and CCl4. We provide a critical comparison of our results with the spectra available in the literature and from online databases and present spectra recommended for use in radiative transfer calculations. Numerous theoretical calculations of infrared absorption cross sections for estimation of global warming potentials (GWP) exist in the literature [1]. In many of

T.J. Wallington et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 174 (2016) 56–64

2.1. Experimental details The apparatus used to measure the spectra has been described previously (Wallington and Japar [14], Pinnock et al. [15]). The apparatus consists of a Mattson Instruments Sirius Fourier transform infrared spectrometer coupled to a 140 l, 2 m long, evacuable Pyrex chamber and a narrow band MCT detector. Internal White type multiple reflection optics provide a path length of 27.4 m. The spectrometer was operated at a spectral resolution of 0.25 cm
1 over the spectral range of 650–3500 cm
1. Frequency calibration was achieved using the interferometer laser with an uncertainty of 70.09 cm
1. Samples of the chloromethanes were obtained from commercial sources at stated purities 499.9% and were subjected to repeated freeze-pump-thaw cycling to remove volatile impurities. Mixtures of the chloromethanes were made up with 700 Torr of air at 295 K and introduced into the chamber and their concentrations were adjusted such that the absorption features were unsaturated.

3. Results 3.1. 1 Experimental results As shown in Fig. 1 the absorption features for CH3Cl, CH2Cl2, CHCl3, and CCl4 scaled linearly with the sample concentrations in the chamber. The linearity of the plots in Fig. 1 indicates that saturation was not a problem in the present work. Absorption spectra for CH3Cl, CH2Cl2, CHCl3, and CCl4 over the range 650–2000 and 650–850 cm
1 are displayed in Figs. 2 and 3, respectively. Uncertainties in the spectra are estimated to arise from the following sources: sample concentration, 72%; path length calibration 71.5%; spectrometer accuracy, 71%; residual baseline offset after subtraction of background, 70.5%; and spectrum noise, 710
20 cm2 molecule
1 (Pinnock et al. [15]). The total uncertainty in the measured absorption cross sections is estimated to be 75%. Integrated absorption cross sections measured in the present and previous studies are given in Tables 1–4. The uncertainties listed in Tables 1–4 for the absorption bands measured experimentally in the present work are 75%; those for previous work are as reported in the literature. As seen from Table 1, with the exception of data from Brown et al. [23] there is good agreement in the integrated absorption cross sections reported for CH3Cl from the

40

-1

2. Methods

Dunning's correlation-consistent aug-cc-pVXZ (X ¼D, T, Q, 5) basis sets [21,22]. Vibrational wavenumbers and infrared intensities, obtained in the harmonic approximation for the title compounds, are provided in Tables S1–S4 in the supplementary information. Results from B3LYP, M06-2X and MP2 calculations on the fluorinated analogs (CH3F, CH2F2, CHF3 and CF4) employing the 6-31G* and aug-cc-pVTZ basis sets are provided in Tables S5–S8 in the supplementary information. Additional anharmonic calculations were carried out employing the aug-cc-pVTZ basis set.

Integrated absorbance (cm )

these studies the choice of quantum chemistry level of theory and basis set size is dictated by routine or by limited computational resources; examples of the commonly employed methods include B3LYP/6-31G* [2–5], B3LYP/631G** [6,7], B3LYP/6-311G** [8], and MP2/6-31G** [5]. Systematic studies show that the abovementioned methods are far from optimal. Jensen [9–12] has benchmarked basis sets suitable for systematically approaching the basis set limit for Hartree-Fock and density functional methods. In short, Jensen's systematic studies show that basis sets of at least triple-zeta quality augmented by diffuse functions are necessary to ensure that the calculated structures, harmonic vibrational frequencies and infrared intensities are reasonably close to the basis set limit results. Galabov et al. [13] investigated various correlated levels of theory (MP2, CISD, CCD, QCISD, CCSD and CCSD(T)) and arrived at the same conclusion: at least triple-zeta basis sets augmented by diffuse functions are needed to “produce plausible accord between theory and experiment”. The present work has three goals: (i) to present the results from a comprehensive experimental and computational study of the IR spectra of CH3Cl, CH2Cl2, CHCl3, and CCl4; (ii) to compare these results with the available literature data and provide recommended spectra for use in radiative transfer calculations and experimental studies; (iii) to provide radiative efficiency and global warming potentials for CH3Cl, CH2Cl2, CHCl3, and CCl4.

57

30

CH Cl CH Cl CHCl CCl

20

10

2.2. Computation details 0

Frozen core MP2 [16] and DFT calculations employing the M06-2X [17], Becke 3 parameter [18], and Lee-YangParr [19] B3LYP hybrid functionals were carried out with the Gaussian 09 program [20]. Calculations were carried out employing the frequently used 6-31G* basis set and

0

20

40

60

80

[CHxCl4-x] (mTorr) Fig. 1. Plots of integrated absorbance versus CHxCl4
x concentration: CH3Cl, circles, 660–780 cm
1; CH2Cl2, triangles, 660–800 cm
1; CHCl3, squares, 725–810 cm
1; CCl4, inverted triangles, 730–825 cm
1.

58

T.J. Wallington et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 174 (2016) 56–64

0.6

Table 1 Integrated absorption cross sections at 295–298 K for CH3Cl.

CH 3Cl

σ (10 -18 cm 2 molecule -1)

0.4

Wavenumber range

Integrated cross section a

Referenceb

0.2 0.0

660–780

0.389 70.020 0.376 (0.386)

This work This work (computational)c Elkins et al. [24] Dickson et al. [25] Barrow and McKean [26] Bera et al. [40] (computational) Sharpe et al. [27] This work This work (computational)c Elkins et al. [24] Dickson et al. [25] Barrow and McKean [26] Bera et al. [40] (computational) Sharpe et al. [27] Brown et al. [28] This work Sharpe et al. [27]

0.405 70.013 0.388 70.003 0.350 0.352

CH 2Cl 2

0.5

0.0

1290–1610

2

CHCl 3

1

0.409 70.011 0.3277 0.016 0.376 (0.355) 0.325 70.014 0.319 70.081 0.312 0.440

0 4

CCl 4

2

697–1377

0 700

900

1100

1300

1500

1700

1900

-1

Wavenumber (cm ) Fig. 2. Absorption spectra for CH3Cl, CH2Cl2, CHCl3, and CCl4 at 600-2000 cm
1 measured in 700 Torr of air at 295 K.

Fig. 3. Absorption spectra for CH3Cl, CH2Cl2, CHCl3, and CCl4 at 650– 850 cm
1 measured in 700 Torr of air at 295 K.

experimental measurements in the present work, Elkins and Kagann [24], Dickson et al. [25], Barrow and McKean [26], and the spectrum in the online PNNL database from Sharpe et al. [27]. For reasons that are unclear the results reported by Brown et al. [28] are approximately 25% lower than from our measurements. It is clear that the IR

0.342 70.010 0.400 0.5447 0.027 0.598 70.018

: Units of 10
17 cm2 molecule
1 cm
1. : Experimental studies except where noted otherwise. c : Results from MP2(Full)/aug-cc-pVTZ calculations, anharmonic values in brackets. a

b

spectrum of CH3Cl is well established. With the exception of the results from Brown et al. [28], there is no obvious reason to exclude any of the previous studies in making a recommendation for the integrated absorption cross section. Taking an average of the results from Elkins et al. [24], Dickson et al. [25], Barrow and McKean [26], Sharpe et al. [27], and the present work gives our recommended integrated absorption over the range 660–780 cm
1 of 3.89  10
18 cm
1 molecule
1. For CH2Cl2 we can compare our spectrum with those available in the PNNL (Sharpe et al. [27]) and NIST (Chu et al. [29]) online databases. As shown in Table 2, over the wavenumber range 650–3075 cm
1 the PNNL spectrum is approximately 7% more intense while the NIST spectrum is approximately 10% more intense than that measured in the present work. We estimate total uncertainties of 75% in the present work while the quoted uncertainties for spectra in the PNNL and NIST databases are 3% and 2.1%, respectively. The integrated absorption cross sections in the Ford and PNNL spectra are indistinguishable within the combined uncertainties. There is an approximately 10% difference between the integrated absorption cross sections in the Ford and NIST spectra which is just outside the combined uncertainty ranges. It is not obvious that there are any systematic errors in either our study or that from NIST. We choose to take an average from the three studies to derive a recommended integrated absorption over the range 650–800 cm
1 of 2.16  10
17 cm
1 molecule
1. For CHCl3 inspection of Table 3 shows there is a significant range in the reported cross sections. The strength of absorption in the 720–810 cm
1 region reported by McPheat and Duxbury [30] is approximately 35% greater than that reported by Kim and King [31] and Tanabe and

T.J. Wallington et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 174 (2016) 56–64

Table 2 Integrated absorption cross sections at 295–298 K for CH2Cl2.

59

Table 3 Integrated absorption cross sections at 295–298 K for CHCl3.

Wavenumber range

Integrated cross sectiona

Referenceb

Wavenumber range

Integrated cross sectiona

Referenceb

650–800

2.08 70.11 2.15 (2.18)

This work This work (computational) c Sharpe et al. [27] Chu et al. [29] This work This work (computational) c Sharpe et al. [27] Chu et al. [29] This work This work (computational) c Sharpe et al. [27] Chu et al. [29] This work This work (computational) Sharpe et al. [27] Chu et al. [29] Bera et al. [40], computational This work (computational) c

720–810

4.13 70.21 4.38 (4.40)

This work This work (computational)c McPheat and Duxbury [30] Kim and King [31] Tanabe and Saëki [32] Sharpe et al. [27] Chu et al. [29] This work This work (computational)c McPheat and Duxbury [30] Kim and King [31] Tanabe and Saëki [32] Sharpe et al. [27] Chu et al. [29] This work This work (computational)c McPheat and Duxbury [30] Kim and King [31] Tanabe and Saëki [32] Bera et al. [40] (computational) Sharpe et al. [27] Chu et al. [29]

1240–1300

2920–3075

650–3075

250–3300

2.1670.07 2.23 70.05 0.4997 0.025 0.636 (0.545) 0.5177 0.16 0.521 70.11 0.107 70.054 0.091 (0.109) 0.113 70.036 0.114 70.024 2.69 70.13 2.90 (2.91) 2.87 70.09 2.95 70.06 3.30 2.90 (2.92)

5.050 7 0.146

1175–1255

3.63 7 0.04 3.75 7 0.30 4.21 70.13 4.39 7 0.10 0.561 70.028 0.699 (0.635) 0.557 70.023b

650–1255

0.503 7 0.017 0.5317 0.043 0.593 7 0.18 0.593 7 0.13 4.777 0.24 5.13 (5.09) 5.83 7 0.10 4.21 70.06 4.357 0.35 5.46

: Units of 10
17 cm2 molecule
1 cm
1. : Experimental studies except where noted otherwise. c : Results from MP2(Full)/aug-cc-pVTZ calculations, anharmonic values in brackets. a

4.92 7 0.15 5.09 7 0.11

b

: Units of 10
17 cm2 molecule
1 cm
1. : Experimental studies except where noted otherwise. c : Results from MP2(Full)/aug-cc-pVTZ calculations, anharmonic values in brackets. a

Saëki [32] and approximately 25% greater than measured in the present work. However, the intensities reported for the weaker absorption band at 1175–1255 cm
1 in the four studies are consistent within the experimental uncertainties. The measurements reported by Sharpe et al. [27] and Chu et al. [29] are consistent with those in the present work. To better understand the discrepancy between our measurement and that of McPheat and Duxbury [30] we attempted to obtain an electronic version of the spectrum from Prof. Duxbury, but learned that the spectra recorded at 298 K are no longer available. However, the spectrum at 253 K is available in the supporting information in Hodnebrog et al. [1] listed under spectra from Highwood and Shine [33]. Fig. 4 shows a comparison of the spectrum of CHCl3 from McPheat and Duxbury [30] and that from the present work. There is considerably more noise in the spectrum from McPheat and Duxbury [30]. To our surprise in plotting the two spectra it is clear that there is little, or no, significant difference in the intensity of absorption in the strong band at 720– 810 cm
1 but there is a significant difference in the results for the weaker band at 1175–1255 cm
1. Integrating the absorption in the 720–810 cm
1 region from McPheat and Duxbury [30] gives a value of 4.38  10
17 cm
1 molecule
1 which is consistent with our results. We suggest that there was an error in integrating the absorption band by McPheat and Duxbury [30]. Taking an average of the results from Kim and King [31], Tanabe and Saëki [32], reevaluated data from McPheat and Duxbury [30], Sharpe et al. [27], Chu et al. [29], and the present work gives a

b

Table 4 Integrated absorption cross sections at 295–298 K for CCl4. Wavenumber range

Integrated cross sectiona

Referenceb

730–825

6.22 7 0.31 6.65

This work This work (computational)c Nemtchinov and Varanasi [34] Orlando et al. [35] Tanabe and Saëki [32] Lindsay and Schatz [36] Bera et al. [40], computational Sharpe et al. [27] Chu et al. [29] Fisher et al. [37] This work Orlando et al. [35] Zander et al. [38] Sharpe et al. [27] Chu et al. [29] This work Orlando et al. [35] Massie et al. [39] Sharpe et al. [27] Chu et al. [29]

6.295 7 0.105 5.89 7 0.29 5.357 1.00 6.487 0.52 7.37

616–934 773–802

786–806

a b c

6.39 7 0.19 6.717 0.13 4.45 4.95 7 0.25 4.62 7 0.23 3.68 5.107 0.15 5.40 7 0.11 4.39 7 0.22 4.12 70.21 3.59 4.5170.14 4.78 7 0.10

: Units of 10
17 cm2 molecule
1 cm
1. : Experimental studies except where noted otherwise. : Results from MP2(Full)/aug-cc-pVTZ calculations.

60

T.J. Wallington et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 174 (2016) 56–64

3.2. Computational results

Fig. 4. Absorption spectrum for CHCl3 at 1200–1240 cm
1 (top panel) and 730–800 cm
1 (bottom panel) reported by McPheat and Duxbury [30] (gray) and in the present work (black).

recommended integrated absorption over 720–810 cm
1 of 4.08  10
18 cm
1 molecule
1. For CCl4 the integrated absorption cross sections measured in the present work are in agreement within the experimental uncertainties with the results reported by Nemtchinov and Varanasi [34], Orlando et al. [35], Tanabe and Saëki [32], and Lindsay and Schatz [36], Sharpe et al. [27], and Chu et al. [29], but are approximately 20–30% greater than those reported by Fisher et al. [37], Zander et al. [38] and Massie et al. [39]. The measurement techniques and calibration procedures employed in the absorption cross section measurements reported by Fisher et al. [37], Zander et al. [38] and Massie et al. [39] are not well documented. In the title of their paper Massie et al. (1985) state that their measurements are "approximate" and in the text they state "some of the measurements may not be in the linear region". However, for the pressure (0.34 Torr) and path length (5 cm) employed by Massie et al. [39] the peak transmission by CCl4 at 795.5 cm
1 would only be approximately 23% and hence the measurement should be in the linear region. The integrated absorption cross sections reported by Massie et al. [39] for other gases are generally in good agreement with literature studies [1]. Hence, it is difficult to offer an explanation for the discrepancy between their results and those from the present work. Fisher et al. [37] report IR absorption cross section data for several halocarbons, most of which are in good agreement with other literature measurements [1] and it is difficult to offer an explanation of the discrepancy with the results from the present study. Taking an average of the results from Nemtchinov and Varanasi [34], Orlando et al. [35], Sharpe et al. [27], and Chu et al. [29], and the present work gives a recommended integrated absorption over the range 730–825 cm
1 of 6.30  10
17 cm
1 molecule
1.

Tables S1–S4 in the supplemental information illustrate the variation in calculated harmonic vibrational frequencies and the associated integrated intensities. In general, the results concord with the findings of Jensen [9–12] – basis sets of at least triple-zeta quality augmented by diffuse functions should be employed to obtain reliable harmonic vibrational frequencies and infrared intensities. It is also obvious from Tables S1–S4 that the B3LYP functional fails to describe the C–Cl bonds correctly; the predicted wavenumbers of the C–Cl stretching modes are all lower than the observations. The same deficiency is also observed for C–F stretching vibrations in CH3F, CH2F2, CHF3 and CF4, Tables S5–S8, and is attributed to an inadequate treatment of the short-range exchangecorrelation energy (see e.g. Zhao and Truhlar [41]). The B3LYP functional is therefore not the best choice for estimating vibrational frequencies and infrared intensities of halocarbons. Only seldomly are integration and convergence criteria considered or discussed in relation to calculated vibrational frequencies and infrared intensities intended for calculation of radiative forcing and GWP. The same applies to the effect of including the inner shells in the MP2 correlation calculations. The differences between frequencies and intensities obtained using “standard” criteria for structure and SCF convergence and grid-size used for numerical integrations, and more tight criteria and finer grids, are not always negligible. Table S9 in the supplemental information illustrates this issue for B3LYP/aug-ccpVTZ and M06-2X/aug-cc-pVTZ calculations on CH2Cl2. It can be concluded that a standard grid will be adequate for most B3LYP calculations and that ultrafine grids should be used in M06-2X calculations on halocarbons; the associated increased computational cost is very moderate. Inclusion of all electrons in the MP2 correlation calculation results in relatively moderate changes in the harmonic vibrational frequencies and intensities with only a small increase in computational cost for the systems studied. Table S10 in the supplemental information compares the results from MP2(Frozen Core) and MP2(Full) calculations on the four chloromethanes. We recommend MP2(Full)/ aug-cc-pVTZ calculations for estimating vibrational frequencies and intensities of halocarbons whenever possible, and M06-2X/aug-cc-pVTZ calculations when MP2 calculations are not feasible. Assuming a sufficient basis set size, the calculated harmonic vibrational frequencies will ideally be higher than the observed fundamental frequencies. This is due to the neglect of anharmonicity and is commonly compensated for by “scaling”. Table 5 compares observed fundamental vibrational frequencies for the chloromethanes with the results from the harmonic calculations including scaling of the harmonic vibrational frequencies. It can be seen that the scaled harmonic vibrational frequencies in general are in quite good agreement with the observed fundamental frequencies – even for the small basis set calculations (6-31G*); the B3LYP predictions of the C–Cl stretching modes are, however, always too low (the same

T.J. Wallington et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 174 (2016) 56–64

Table 5 Calculated harmonic, fundamental and anharmonic vibrational frequencies (/cm
1) and integrated Intensities (/cm2 cm
1 molecule
1) of CHCl3, CH2Cl2, CHCl3 and CCl4. Results from B3LYP/6-31G*, B3LYP/aug-ccpVTZ, M06-2X/aug-cc-pVTZ and MP2(Full)/aug-cc-pVTZ calculations. Compound and observed fundamental frequencya CH3Cl 3041 (E)

2966 (A1)

1455 (E)

1355 (A1)

1015 (E)

732 (A1)

CH2Cl2 3040 (B1)

2999 (A1)

1467 (A1)

1268 (B2)

1153 (A2)

898 (B1)

758 (B2)

717 (A1)

282 (B1)

Calculated harmonic frequency and intensityb

Calculated fundamental frequencyc

Calculated anharmonic frequency and intensity

3195 3166 3206 3220 3096 3071 3108 3130 1510 1483 1492 1520 1414 1377 1385 1405 1045 1026 1033 1051 721 710 755 774

(25) (19) (7.5) (7.7) (38) (39) (32) (37) (20) (20) (20) (20) (32) (21) (18) (18) (14) (9.0) (6.7) (7.1) (48) (44) (44) (38)

3072 3065 3039 3114 2976 2973 2946 3027 1452 1436 1414 1470 1359 1333 1311 1358 1005 993 979 1016 693 688 716 748

3045 3022 3041 3084 3011 2978 2977 3046 1473 1443 1452 1475 1382 1345 1350 1373 1029 1009 1014 1031 705 693 735 759

(29) (22) (10) (17) (41) (43) (31) (39) (21) (20) (21) (20) (27) (17) (15) (16) (13) (7.0) (10) (7.1) (50) (47) (46) (39)

3224 3198 3225 3223 3145 3122 3150 3181 1490 1460 1474 1503 1321 1285 1298 1313 1196 1170 1187 1208 914 901 908 924 728 715 776 805 705 698 734 747 283 277 284 291

(0.78) (0.50) (1.5) (1.1) (15) (8.7) (8.4) (7.9) (0.13) (0.14) (0.03) (0.20) (91) (65) (68) (64) (0) (0) (0) (0) (3.5) (2.0) (2.1) (2.2) (267) (250) (226) (198) (25) (19) (22) (17) (1.0) (0.66) (0.89) (0.70)

3099 3096 3057 3126 3023 3022 2986 3076 1432 1414 1398 1453 1270 1244 1230 1270 1150 1133 1125 1168 879 872 861 894 700 692 735 779 677 675 696 722 272 269 269 282

3072 3054 3003 3097 3026 2999 2949 3067 1454 1424 1439 1459 1296 1260 1271 1285 1172 1145 1162 1181 903 889 902 911 710 696 745 786 693 686 710 735 281 274 282 288

(2.8) (0.01) (0.84) (0.18) (19) (12) (68) (11) (0.18) (0.00) (0.26) (0.25) (81) (65) (55) (55) (0) (0) (0) (0) (3.2) (2.0) (1.7) (2.1) (271) (215) (226) (201) (26) (19) (21) (17) (0.92) (0.72) (0.81) (0.67)

61

Table 5 (continued ) Compound and observed fundamental frequencya CHCl3 3034 (A1)

1220 (E)

774 (E)

680 (A1)

366 (A1)

260 (E)

CCl4 776 (T2)

459 (A1)

314 (T2)

217 (E)

Calculated harmonic frequency and intensityb

Calculated fundamental frequencyc

Calculated anharmonic frequency and intensity

3201 3179 3203 3230 1265 1232 1250 1269 737 724 789 814 668 662 681 699 367 361 371 380 261 255 262 269

(0.07) (2.4) (3.2) (2.9) (103) (72) (72) (69) (624) (540) (481) (438) (12) (6.4) (7.0) (5.2) (0.71) (0.32) (0.60) (0.32) (0.06) (0.01) (0.25) (0.16)

3077 3078 3036 3123 1216 1193 1185 1227 708 701 748 787 642 641 645 676 353 349 351 367 251 247 248 260

3057 3037 2921 3098 1239 1208 1230 1243 721 710 757 797 660 654 672 690 364 357 368 376 258 253 261 268

745 732 807 821 451 447 474 478 316 309 319 326 218 213 221 226

(1012) (804) (717) (665) (0) (0) (0) (0) (0.71) (2.2) (0.13) (0.21) (0) (0) (0) (0)

716 708 765 794 434 433 450 463 303 299 303 315 210 206 209 219

(1.0) (0.57) (3.3) (1.1) (94) (65) (64) (63) (626) (545) (488) (440) (13) (6.8) (7.7) (5.6) (0.68) (0.31) (0.51) (0.32) (0.03) (0.02) (0.12) (0.18)

a

From NIST [43]. Results are listed in the order: B3LYP/6-31G*, B3LYP/aug-cc-pVTZ, M06-2X/ aug-cc-pVTZ, MP2/aug-cc-pVTZ. c Scaling constants: B3LYP/aug-cc-pVTZ, 0.968 [43]; M06-2X/aug-ccpVTZ, 0.948 (Alecu et al. [44]); MP2(Full)/aug-cc-pVTZ, 0.967 [43]. b

applies to the C–F stretching vibrations in the fluoromethanes, see Tables S5 – S8). The anharmonic contributions to the calculated integrated harmonic intensities, however, cannot readily be obtained through simple scaling of the transition moments or dipole moment derivatives in the same manner. Only recently have reliable procedures been implemented in quantum chemistry packages [42]. However, anharmonic vibrational frequency calculations are by no means yet a standard black box procedure (it should be noted that the anharmonic calculations by necessity were carried out employing very tight convergence criteria and a superfine spherical grid with 96 radial shells having 36θ and 72φ points in each shell). Table 3 includes the results from

62

T.J. Wallington et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 174 (2016) 56–64

anharmonicity calculations for CH3Cl, CH2Cl2 and CHCl3; similar calculations on CCl4 were unsuccessful due to issues related to vibrational degeneracy.

4. Radiative efficiency, GWP, and GTP calculations

Radiative Forcing -3 -2 -18 -1 -1 (10 W m (10 cm molecule )

To provide spectra for use in radiative efficiency calculations we scaled those reported in the PNNL database

3.0 2.5 2.0 1.5 1.0 0.5 0.0 0

500

1000

1500

2000

-1

Wavenumber (cm )

8

2.4

6

1.8

4

1.2

2

0.6

0

0.0 660

680

700

720

740

760

780

800

820

cm molecule )

3.0 Pinnock curve CCl4 Exp. abs. cross-sec. B3LYP/aTZ B3LYP/aTZ scaled

Radiative Forcing /10 W m cm (10

10

Absorption cross-section (base e) /10

cm molecule

Fig. 5. Radiative forcing (for a 0–1 ppb increase in mixing ratio) per unit cross section per cm
1 (Hodnebrog et al. [1]). Numerical values are given in the supplementary data.

840

Wavenumber /cm Fig. 6. Radiative forcing (for a 0–1 ppb increase in mixing ratio) per unit cross section per cm
1 (Pinnock curve), experimental absorption crosssection of CCl4, and theoretical rectangular absorption cross sections of CCl4 from B3LYP/aug-cc-pVTZ calculations in the 650–850 cm
1 region.

for CH3Cl, CH2Cl2, CHCl3, and CCl4 by factors of 0.95, 1.00, 0.97, and 0.99, respectively, to reproduce the integrated absorption cross sections recommended above. The resulting recommended spectra are provided in digital form in the Supporting Information. We chose to use spectra from the PNNL database because these spectra cover a wide wavenumber range, have high signal to noise, are free of trace water features, have been accurately wavelength calibrated, are freely available at the PNNL website, and the PNNL dataset includes all four of the title compounds in a consistent format. We use the simple method of Hodnebrog et al. [1] to estimate the radiative efficiency from IR absorption spectra of CH3Cl, CH2Cl2, CHCl3, and CCl4. This method is a revised version of the widely used Pinnock et al. [15] method, where the instantaneous RE can be estimated directly from the absorption spectrum of a gas without the use of a radiative transfer model. The most notable improvements since Pinnock et al. [15] were better representation of clouds and that a more advanced radiative transfer model were used as a basis for the calculations (line-by-line code instead of a narrow band model). A resolution of 1 cm
1 for the spectral bands was used here to calculate the instantaneous radiative forcing of a 1 ppb change in concentration for each compound. The radiative forcing per unit IR absorption used is shown as a function of wavenumber in Fig. 5. Assuming a 10% increase from the instantaneous RE due to stratospheric temperature adjustment [1], we calculate REs for CH3Cl, CH2Cl2, CHCl3, and CCl4 of 0.005, 0.050, 0.127, and 0.186 W m
2 ppb
1, respectively, from the experimental spectra. To account for the fact that the chloromethane gases have a non-uniform horizontal and vertical distribution in the atmosphere, with CCl4 mainly being lost due to photolysis in the stratosphere and the remaining three compounds due to reaction with tropospheric OH, we apply lifetime correction methods according to Hodnebrog et al. [1]. The resulting REs for CH3Cl, CH2Cl2, CHCl3, and CCl4, including both stratospheric temperature adjustment and lifetime correction, and assuming atmospheric lifetimes from WMO [45] of 1.0, 0.4, 0.4, and 26 years, respectively, are 0.004, 0.028, 0.070, and 0.174 W m
2 ppb
1 (see Table 6). The radiative efficiency of CH3Cl in the fifth assessment report (AR5) of the Intergovernmental Panel on Climate Change (IPCC) [46] is taken from Grossman et al. [47] who derived an instantaneous RE of 0.005 W m
2 ppb
1, in agreement with our study. The RE of CH2Cl2 used in IPCC AR5, and previous assessments, is based on a personal communication from Fisher. Their value of

Table 6 Lifetimes, radiative efficiencies (RE), and direct global warming potentials (GWP) and global temperature-change potentials (GTPs). Gas

CH3Cl CH2Cl2 CHCl3 CCl4 a

τ (yr)a

1.0 0.4 0.4 26

RE (W m
2 ppb
1)

GWP

GTP

IPCC AR5 [46]

This study

20-yr

100-yr

20-yr

50-yr

100-yr

0.01 0.03 0.08 0.17

0.004 0.028 0.070 0.174

17 29 54 3570

5 8 15 1775

6 8 17 3356

1 1 2 1612

1 1 2 490

Lifetimes are from WMO [45].

T.J. Wallington et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 174 (2016) 56–64

0.03 W m
2 ppb
1 is in agreement with the lifetimecorrected RE derived here. For CHCl3, the RE in IPCC AR5 was calculated using the same method as in this study (see Hodnebrog et al. [1]), but based only on the absorption cross-section from McPheat and Duxbury [30]. Their slightly higher value of 0.08 W m
2 ppb
1 is simply due to the higher integrated absorption cross-section in the spectrum of McPheat and Duxbury [30] compared to our recommended spectrum for CHCl3 (Table 1). Our RE for CCl4 is in excellent agreement with the value of 0.17 W m
2 ppb
1 used in IPCC AR5 based on the spectrum of Nemtchinov and Varanasi [34]. Using the lifetime-corrected RE for each compound, we present Global Warming Potentials (GWP) and Global Temperature-change Potentials (GTP) for various time horizons (Table 6). GWPs and GTPs have been calculated relative to CO2, using the method presented in IPCC AR5 [46]. The results show that only CCl4 has significant GWP and GTP values, while the three other compounds have a negligible impact on radiative forcing of climate change, mainly due to their relatively short lifetime, but also due to weaker radiative efficiencies. Our study provides results for two compounds (CH3Cl and CH2Cl2) where experimental absorption cross sections have not been available [1], and confirm the values in IPCC AR5, although with slight changes, for the remaining two compounds (CHCl3 and CCl4). The spectra for CH3Cl, CH2Cl2, CHCl3, and CCl4 recommended for use in radiative efficiency calculations are given in the supplementary data. The theoretical infrared absorption cross sections of CH3Cl, CH2Cl2, CHCl3 and CCl4 were employed for estimating their radiative efficiencies. In these calculations the band shapes were approximated by rectangular functions of uniform widths; two widths of respectively 11 and 21 cm
1 were employed to test the influence of bandwidth on the calculated RE. Tables S10–S14 in the supplemental information compares the various theoretical integrated absorption cross sections, instantaneous radiative efficiencies and derived GWP values to the results based on experimental data. Three conclusions can be drawn from the results listed in these Tables. First, the theoretical integrated absorption cross sections, S, are generally larger than the experimental values. Second, the B3LYP results for S show the largest deviations whereas the MP2 results show best agreement with experiment closely followed by the M06-2X results. Third, the instantaneous REs are generally insensitive to the bandwidth employed in the calculations. In spite of the poor performance of B3LYP for calculating vibrational frequencies and intensities, the B3LYP calculations employing the small 6-31G* basis set often result in the better agreement with the “experimental” radiative efficiency and GWP – one gets the right result, but for the wrong reasons. Fig. 6 illustrates how the error in the calculated absorption cross-section for CCl4 is balanced by the error in the calculated vibrational frequency that positions the band in the region where CO2 completely dominates the radiative properties of the atmosphere (Table S4).

63

5. Conclusions The data presented in the present paper resolve discrepancies in the literature concerning the IR spectra of chloromethanes. The existing data are reviewed and recommended spectra are provided (see supplementary data). The first three members of the series (CH3Cl, CH2Cl2, CHCl3) have negligible global warming potentials and do not contribute to radiative forcing of climate change. CCl4 has a radiative efficiency of 0.174 W m
2 ppb
1. The global average tropospheric concentration of CCl4 in 2014 was approximately 85 ppt and decreased at a rate of approximately 1.2 ppt yr
1 over the past decade [48] reflecting decreased emissions of this compound which is controlled under the Montreal Protocol. Combining the 2014 atmospheric concentration with radiative efficiency provides an estimate of 0.015 W m
2 for the direct radiative forcing attributable to CCl4. This can be compared to the radiative forcing of 1.68 W m
2 estimated for CO2 in 2011 [49]. CCl4 makes a small but non-negligible contribution to the direct radiative forcing of climate change.

Acknowledgments We thank Jan Fuglestvedt, Keith Shine, George Marston, and Gunnar Myhre for helpful discussions.

Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j. jqsrt.2016.01.029.

References [1] Hodnebrog Ø, Etminan M, Fuglestvedt JS, Marston G, Myhre G, Nielsen CJ, Shine KP, Wallington TJ. Global warming potentials and radiative efficiencies of halocarbons and related compounds: a comprehensive review. Rev Geophys 2013;51:300–78. [2] Blowers P, Moline DM, Tetrault KF, Wheler RR, Tuchawena SL. Prediction of radiative forcing values for hydrofluoroethers using density functional theory methods. J Geophys Res 2007;112: D15108. http://dx.doi.org/10.1029/206JD008098. [3] Blowers P, Moline DM, Tetrault KF, Wheler RR, Tuchawena SL. Global warming potentials of hydrofluoroethers. Environ Sci Technol 2008;42:1301–7. [4] Blowers P, Tetrault KF, Trujillo-Morehead Y. Theor Chem Acc 2008;119:369–81. http://dx.doi.org/10.107/s00214-007-0394-3. [5] Bravo I, Aranda A, Hurley MD, Marston G, Nutt DR, Shine KP, Smith K, Wallington TJ. Infrared absorption spectra, radiative efficiencies, and global warming potentials of perfluorocarbons: comparison between experiment and theory. J Geophys Res 2010;115: D24317. http://dx.doi.org/10.1029/2010JD014771. [6] Bravo I, Marston G, Nutt DR, Shine KP. Radiative efficiencies and global warming potentials using theoretically determined absorption cross-sections for several hydrofluoroethers (HFEs) and hydrofluoropolyethers (HFPEs). J Quant Spectrosc Radiat Transf 2011;112: 1967–77. [7] Bravo I, Diaz-de-Mera Y, Aranda A, Moreno E, Nutt DR, Marston G. Radiative efficiencies for fluorinated esters: indirect global warming potentials of hydrofluoroethers. Phys Chem Chem Phys 2011;13: 17185–93. [8] Blowers P, Kollingshead K. Estimations of global warming potentials from computational chemistry calculations for CH2F2 and other

64

[9] [10]

[11]

[12]

[13]

[14]

[15]

[16] [17]

[18] [19]

[20]

[21]

[22]

[23]

[24]

[25]

[26] [27]

[28]

[29]

[30]

[31]

T.J. Wallington et al. / Journal of Quantitative Spectroscopy & Radiative Transfer 174 (2016) 56–64 fluorinated methyl species verified by comparison to experiment. J Phys Chem A 2009;113:5942–50. Jensen F. Polarization consistent basis sets: principles. J Chem Phys 2001;115:9113–25. http://dx.doi.org/10.1063/1.1413524. Jensen F. Polarization consistent basis sets. II. Estimating the Kohn– Sham basis set limit. J Chem Phys 2002;116:7372–9. http://dx.doi.org/ 10.1063/1.1465405. Jensen F. Polarization consistent basis sets. III. The importance of diffuse functions. J Chem Phys 2002;117:9234–40. http://dx.doi.org/ 10.1063/1.1515484. Jensen F. Polarization consistent basis sets. IV. The basis set convergence of equilibrium geometries, harmonic vibrational frequencies, and intensities. J Chem Phys 2003;118:2459–63 doi:.1063/1.1535905. Galabov B, Yamaguchi Y, Remington RB, Schafer III HF. High level ab initio quantum mechanicl prediction of infrared intensities. J Phys Chem A 2002;106:819–32. Wallington TJ, Japar SM. Fourier transform infrared studies of the reaction of HONO with HNO3, NO3, and N2O5 at 295 K. J Atmos Chem 1989;9:399–409. Pinnock S, Shine KP, Smyth TJ, Hurley MD, Wallington TJ. Radiative forcing of climate by hydrochlorofluorocarbons and hydrofluorocarbons. J Geophys Res 1995;100:23227–38. Møller C, Plesset MS. Note on an approximate treatment for manyelectron systems. Phys Rev 1934;46:618. Zhao H, Truhlar DG. The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: two new functionals and systematic testing of four M06-class functionals and 12 other functionals. Theor Chem Acc 2008;120:215–41. Becke AD. Density-functional thermochemistry 3. The role of exact exchange. J Chem Phys 1993;98:5648. Lee C, Yang W, Parr RG. Development of the colle-salvetty correlation-energy formula into a functional of the electron-density. Phys Rev B 1988;37:785. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, Scalmani G, Barone V, Mennucci B, Petersson GA. et al. Gaussian, Inc., Wallingford CT: 2009. Dunning Jr. TH. Gaussian-basis sets for use in correlated molecular calculations. 1. The atoms boron through neon and hydrogen. J Chem Phys 1989;90:1007. Kendall RA, Dunning Jr. TH, Harrison RJ. Electron-affinities of the 1st-row atoms revisited-systematic basis-sets and wave-functions. J Chem Phys 1992;96:6796. Brown LR, Farmer CB, Rinsland CP, Toth RA. Molecular line parameters for the atmospheric trace molecule spectroscopy experiment. Appl Opt 1987;26:5154–82. Elkins JW, Kagann RH, Sams RL. Infrared band strengths for methyl chloride in the range of atmospheric interest. J Mol Spect 1984;105: 480–90. Dickson AD, Mills IM, Crawford B. Vibrational intensities VIII: CH3and CD3- chloride, bromide and iodide. J Chem Phys 1957;27: 445–55. Barrow GM, McKean DC. The intensities of absorption bands in the methyl halides. Proc R Soc A 1952;213:27–41. Sharpe SW, Johnson TJ, Sams RL, Chu PM, Rhoderick GC, Johnson PA. Gas-phase databases for quantitative infrared spectroscopy. Appl Spectrosc 2004;58:1452–61. Brown LR, Farmer CB, Rinsland CP, Toth RA. Molecular line parameters for the atmospheric trace molecule spectroscopy experiment. Appl Opt 1987;26:5154–82. Chu PM, Guenther FR, Rhoderick GC, Lafferty WJ. The NIST quantitative infrared database. J Res Natl Inst Stand Technol 1999;104: 1–23. McPheat R, Duxbury G. Infrared absorption cross-sections and integrated absorption intensities of chloroform and fluoroform vapor. J Quant Spectro Radiat Transf 2000;66:153–67. Kim K, King WT. Infrared intensities in chloroform. J Chem Phys 1984;80:978–82.

[32] Tanabe K, Saeki S. Calculation of infrared band intensities of various chlorinated methanes. Spectrochim Acta 1970;26A:1469–79. [33] Highwood EJ, Shine KP. Radiative forcing and global warming potentials of 11 halogenated compounds. J Quant Spectrosc Radiat Transf 2000;66:169–83. [34] Nemtchinov V, Varanasi P. Thermal infrared absorption crosssections of CCl4 needed for atmospheric remote sensing. J Quant Spectrosc Radiat Transf 2003;82:473–81. [35] Orlando JJ, Tyndall GS, Huang A, Calvert JG. Temperature dependence of the infrared absorption cross sections of carbon tetrachloride. Geophys Res Lett 1992;19:1005–8. [36] Lindsay LP, Schatz PN. Absolute infrared intensities in CHBr3(l), CHBr3(g) and CCl4(g). Spectrochim Acta 1964;20A:1421–9. [37] Fisher DA, Hales CH, Wang W-C, Ko MKW, Sze ND. Model calculations of the relative effects of CFCs and their replacements on global warming. Nature 1990;344:513–6. [38] Zander R, Rinsland CP, Farmer CB, Norton RH. Infrared spectroscopic measurements of halogenated source gases in the stratosphere with the ATMOS instrument. J Geophys Res 1987;92:9836–50. [39] Massie ST, Goldman A, Murcray DG, Gille JC. Approximate absorption cross sections of F12, F11, ClONO2, N2O5, HNO3, CCl4, CF4, F21, F113, F114, and HNO4. Appl Opt 1985;19:3426–7. [40] Bera PP, Francisco JS, Lee TJ. Identifying the molecular origin of global warming. J Phys Chem A 2009;113:12694–9. [41] Zhao H, Truhlar DG. The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: two new functionals and systematic testing of four M06-class functionals and 12 other functionals. Theor Chem Acc. 2008;120:215–41. [42] Bloino J, Barone V. A second-order perturbation theory route to vibrational averages and transition properties of molecules: general formulation and application to infrared and vibrational circular dichroism spectroscopies. J Chem Phys 2012;136:124108. [43] NIST, 2013. Computational Chemistry Comparison and Benchmark Database, NIST Standard Reference Database Number 101 Release 16a, August 2013, Editor: Russell D. Johnson III 〈http://cccbdb.nist. gov/〉. [44] Alecu IM, Zheng J, Zhao Y, Truhlar DG. Computational thermochemistry: scale factor databases and scale factors for vibrational frequencies obtained from electronic model chemistries. J Chem Theory Comput 2010;6:2872–87. [45] WMO. Scientific Assessment of Ozone Depletion: 2010, Global Ozone Research and Monitoring, Project-Report No. 50. Geneva, Switzerland: World Meteorological Organization; 572. [46] Myhre G, Shindell D, Bréon F-M, Collins W, Fuglestvedt J, Huang J, Koch D, Lamarque J-F, Lee D, Mendoza B, Nakajima T, Robock A, Stephens G, Takemura T, Zhang H. Anthropogenic and natural radiative forcing. In: Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM, editors. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press; 2013. [47] Grossman AS, Grant KE, Blass WE, Wuebbles DJ. Radiative forcing calculations for CH3Cl and CH3Br. J Geophys Res Atmos 1997;102 (D12):13651–6. [48] NOAA, 2014, Earth Systems Research Laboratory, 〈http://www.esrl. noaa.gov/gmd/hats/insitu/cats/conc/brwccl4.html〉, downloaded 30th October, 2014. [49] IPCC. Climate change 2013: the physical science basis. In: Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM, editors. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, United Kingdom: Cambridge University Press; 2013.