Stability and infrared spectra of mono-, di-, and trichloromethanol

Article abstract of DOI:10.1016/S0009-2614(00)00384-5

CH₂ClOH, CHCl₂OH, and CCl₃OH were prepared by UV irradiation of CH₃OH/Cl₂/N₂ gas mixtures. Absorption cross-sections (base e) of σ(CH₂ClOH) at 1093 cm^-1=6.16×10^-1

Full text of DOI:10.1016/S0009-2614(00)00384-5

1
2 May 2000
Chemical Physics Letters 322 Ž2000. 97–102
www.elsevier.nlrlocatercplett
Stability and infrared spectra of mono-, di-, and trichloromethanol
T.J. Wallington a,), W.F. Schneider , I. Barnes , K.H. Becker , J. Sehested ,
a
b
b
c
d
O.J. Nielsen
a
Ford Motor Company, Ford Research Laboratory, 20000 Rotunda DriÕe, Mail Drop SRL-3083, P.O. Box 2053, Dearborn,
MI 48121-2053, USA
b
Bergische UniÕersit a¨ t Wuppertal, Physikalische Chemie - FB9, Gauss-Strasse 20, D-42097 Wuppertal, Germany
c
Haldor Topsøe ArS, NymølleÕej 55, DK-2800 Lyngby, Denmark
d
UniÕersity of Copenhagen, Department of Chemistry, KL5, UniÕersitetsparken 5, 2100 Copenhagen Ø, Denmark
Received 24 February 2000; in final form 13 March 2000
Abstract
CH ClOH, CHCl OH, and CCl OH were prepared by UV irradiation of CH OHrCl rN gas mixtures. Absorption
2
2
3
3
2
2
y
1
y19
y
1
y18
cross-sections Žbase e. of s ŽCH ClOH. at 1093 cm s6.16=10 , s ŽCHCl OH. at 1109 cm s1.22=10 , and
2
2
y
1
y19
2
y1
s ŽCCl OH. at 1119 cm s8.96=10
cm molecule
were determined. The chlorinated methanols decayed with
3
first-order kinetics to HCl and the corresponding carbonyl compound. The decay rates increased with increased contact of
y
2
the chloromethanols with the reactor walls, indicating that decomposition is heterogeneous. An upper limit of 1.05=10
y
1
s
was established confirm this slow rate of homogeneous decomposition. q 2000 Published by Elsevier Science B.V. All
rights reserved.
1
. Introduction
following Cl or F atom initiated oxidation of the
replacement in question in 700–760 Torr of air are
measured and used to infer the likely atmospheric
oxidation mechanism. In several studies some rather
unusual chemical compounds such as CF O CF ,
Recognition of the adverse effect of chlorofluoro-
carbon ŽCFC. release into the atmosphere w1,2x has
led to an international effort to replace CFCs with
environmentally acceptable alternatives. There has
been a substantial research effort to determine the
atmospheric chemistry and hence environmental im-
pact of the replacements w3x. Smog-chamber systems
3
3
3
CF OH, and CCl OH are formed. Although these
3
3
species are not likely to be of any importance in the
atmosphere, it is important to have an understanding
of their chemical and physical behavior to interpret
accurately the results from smog-chamber experi-
ments.
have been used to provide insight into the atmo-
spheric degradation mechanism of many such com-
pounds w4–6x. Typically, in such studies the products
Recently, Hasson and Smith w7x reported the re-
sults of a product study of the Cl atom initiated
oxidation of several chloroethenes. To explain the
)
Ž
.
rapid formation of CCl O phosgene , Hasson and
Smith w7x postulated a mechanism in which CCl O is
Corresponding author. Fax: q1-313-594-2923; e-mail:
2
twalling@ford.com
2
0
009-2614r00r$ - see front matter q 2000 Published by Elsevier Science B.V. All rights reserved.
PII: S0009-2614Ž00.00384-5

9
8
T.J. Wallington et al.rChemical Physics Letters 322 (2000) 97–102
formed by the rapid Žk s10 sy1. homogeneous
6
surrounded by 22 lamps. Reactant and products were
monitored in situ by FTIR spectroscopy, using spec-
1
decomposition of CCl OH.
3
y1
tral resolutions of either 1 cm ŽWuppertal. or 0.25
CCl OH CCl OqHCl .
1
Ž .
3
2
y1
cm
ŽFord. and analyzing path-lengths of 50 m
ŽWuppertal. or 27 m ŽFord.. All experiments were
performed in 700 Torr total pressure of N diluent at
Hasson and Smith noted that their model was rela-
tively insensitive to the rate for reaction Ž1. and that
2
2
95"2 K. The chloromethanols were prepared by
k1 could be reduced by several orders of magnitude
without significant change in the computed results.
However, even a value of k1 which is ‘several
irradiation of CH OHrCl rN mixtures w8x:
3
2
2
6
y1
Cl
2
qhn 2Cl ,
Ž
Ž
Ž
Ž
Ž
5
6
7
8
9
.
.
.
.
.
.
.
orders of magnitude’ less than 10 s is surprising
given the known stability of analogous halogenated
methanols such as CH ClOH, CHCl OH, and
ClqCH OH CH OHqHCl ,
3
2
2
2
CF OH which have been observed in the gas phase
CH OHqCl
CH ClOHqCl ,
3
2
2
2
and which have upper limits for homogeneous de-
y3
y3
composition of k -1.7=10 w8x, k -9.8=10
CH ClOHqCl CHClOHqHCl ,
2
2
3
y5 y1
w8x, and k -4=10
s
Žhalf-life greater than 5
4
CHClOHqCl2 CHCl OHqCl ,
h. w9x.
2
CHCl OHqCl CCl OHqHCl ,
Ž
Ž
10
11
CH ClOH HCHOqHCl ,
Ž
Ž
Ž
2
3
4
.
.
.
2
2
2
CHCl OH HC
Ž
O
.
ClqHCl ,
CCl OHqCl
CCl OHqCl .
2
2
2
3
CF OH COF qHF .
3
2
The chain is terminated by loss of Cl atoms via
reactions at the chamber wall, by Cl atom recom-
bination, and by association of Cl atoms with other
To investigate the apparent difference in the ho-
mogeneous decomposition rate of CCl OH com-
3
radicals, e.g., ClqCH OH. Initial reaction mixtures
2
pared to CH ClOH and CHCl OH, the three alco-
2
2
consisted of 4–8 mTorr of CH OH and 100–150
3
hols were prepared in smog-chamber experiments
and their decay followed using FTIR spectroscopy.
Infrared ŽIR. spectra and decomposition rates of
mTorr of Cl in 700 Torr of N at 296"3 K.
2
2
CH ClOH and CHCl OH have been reported previ-
2
2
ously w8x. In this work, we report the first study of
3
. Computational details
the IR spectrum of CCl OH and its rate of decompo-
3
sition, and the first absolute IR absorption cross-sec-
tion measurements for CH ClOH and CHCl OH in
the O–H stretching region. The homogeneous de-
Ab initio molecular orbital calculations were per-
2
2
formed using the GAMESS package w12x. The struc-
tures of all species were determined by gradient
optimizations at the all-electron second-order
composition rate of CCl OH was also estimated
3
using ab initio calculations.
UU
Møller–Plesset level ŽMP2. using a 6-31G basis,
and are in good agreement with the Hartree–Fock
structures reported previously w13x. Vibrational fre-
2
. Experimental
quencies were determined by two-sided numerical
differentiation of analytical gradients, and were
scaled by 0.94 to correct for known systematic errors
w14x. To improve the energy estimates, coupled clus-
ter calculations wCCSDŽT., with core electrons frozenx
Experiments were conducted in Wuppertal and at
the Ford Research Laboratory in Dearborn. Experi-
ments in Wuppertal were performed using a 405 l
Pyrex reactor w10x surrounded by 8 super actinic
fluorescent lamps ŽPhilips TL05r40W, 320-l-
UU
were performed at the MP2r6-31G -derived ge-
ometries using Dunning’s correlation-consistent
triple-zeta basis set Žcc-pVTZ, w15x. and the AcesII
code w16x.
4
80 nm, l s360 nm.. At Ford the experiments
max
were performed using a 140 l Pyrex reactor w11x

T.J. Wallington et al.rChemical Physics Letters 322 (2000) 97–102
99
4
. Results
has been proposed as an intermediate in the oxida-
tion of CCl H w18,19x, to the best of our knowledge,
3
Short Ž2–5 s. periods of UV irradiation of
it has not been directly observed previously. For
comparison, the IR spectra have been calculated at
CH OHrCl mixtures in N resulted in the forma-
tion of IR product features attributable to mono- and
dichloromethanol w8x. At longer irradiation times,
these product features were replaced by new features
at 784, 1113, 1311, and 3604 cm which we assign
to CCl OH. Fig. 1 shows IR spectra for the
3
2
2
UU
the MP2r6-31G level, and these results are also
shown in Table 1. The calculated and experimental
spectra are in excellent agreement. Notable observed
features include the O–H stretch mode around 3600
y1
y1
cm , which is slightly red-shifted with increasing
3
chloromethanols obtained by comparing spectra ac-
quired following different periods of UV irradiation
Cl substitution; the C–O–H bend, which is nearly
invariant from mono- to dichloromethanol but red-
y1
and subtracting features attributable to CH OH,
shifted by 80 cm to trichloromethanol; the promi-
3
y1
HCŽO.Cl, and COCl . Measured IR frequencies are
nent C–O stretch mode around 1100 cm , which is
2
listed in Table 1. The IR spectra in Fig. 1 start at 900
cm which was the lowest frequency for which the
slightly blue-shifted with increasing Cl substitution;
y1
and strong C–Cl stretch modes between approxi-
y1
spectrometer is judged to give reliable absolute ab-
sorption cross-sections. The IR frequencies of
CH ClOH and CHCl OH in Table 1 are consistent
mately 700 and 800 cm
.
When reaction mixtures were allowed to stand in
the reaction chambers in the dark the features at-
tributed to CH ClOH, CHCl OH, and CCl OH de-
2
2
with those reported previously w8,17x. While CCl OH
3
2
2
3
cayed and were replaced by HCl and the correspond-
ing carbonyl compounds; HCHO, HCŽO.Cl, or
CCl O. Absolute calibration of the spectra in Fig. 1
2
was achieved by quantifying the absolute increase in
y1
y18
HCHO, HCŽO.Cl Žs Ž1793 cm .s1.63=10 .,
y1
y18
and COCl Žs Ž1832 cm .s1.45=10 ., in the
dark and relating these changes to the corresponding
2
decrease of IR features ascribed to CH ClOH,
2
CHCl OH, and CCl OH, respectively. Calibration
2
3
frequencies were chosen where the chloromethanols
have broad absorption features which were easily
resolved. This approach yielded s ŽCH ClOH. at
2
y1
y19
1
093 cm sŽ6.16"1.23.=10 , s ŽCHCl OH.
2
y1
y18
at 1109 cm s Ž1.22 " 0.35. = 10
,
and
y1
y19
s ŽCCl OH. at 1119 cm sŽ8.96"1.78.=10
3
2
y1
cm molecule . The quoted uncertainty reflects our
estimation of the accuracy of the measurements.
These results can be compared to the previous deter-
y1
minations of s ŽCH ClOH. at 1093 cm sŽ6.2"
2
y19
y1
0
.4.=10
and s ŽCHCl OH. at 1109 cm
s
2
y18
2
y1
Ž1.8"0.2.=10
cm molecule w8x. The value
y1
of s ŽCH ClOH. at 1093 cm
reported herein is
2
indistinguishable from that reported previously w8x.
y1
The value of s ŽCHCl OH. at 1109 cm measured
2
here is approximately 30% lower than that reported
previously w8x reflecting the experimental difficulties
associated with the measurement of the IR spectrum
of a species which has a lifetime of 1–2 min in such
chambers.
Fig. 1. IR spectra of CH ClOH, CHCl OH, and CCl OH.
2
2
3

1
00
T.J. Wallington et al.rChemical Physics Letters 322 (2000) 97–102
Table 1
Observed and calculated vibrational spectra Žcmy1 ., calculated relative intensities Žarbitrary units., and approximate assignments for the
a
chloromethanols
CH ClOH
CHCl OH
CCl OH
2
2
3
Exptl. Calc. Rel. int. Approx. desc. Exptl. Calc. Rel. int. Approx. desc.
Exptl. Calc. Rel. int. Approx. desc.
3
4
694
941
43
32
25
50
61
11
ClCO bend
OH torsion
C–Cl str
CH2 rock
C–O str.
276
316
438
458
673
0
15
6
32
9
CCl2 scissor
CCl2 twist
CCl2 rock
OH torsion
CCl2 sym str
CCl2 asym str
C–O str
CH sym rock
CH asym rock
COH bend
C–H str
231
247
333
344
392
417
522
3
0
26
0
2
15
3
CCl3 d bend
CCl3 d bend
OH torsion
CCl3 s bend
OH rock
6
9
97
60
1
1
1
1
083
176
318
374
1075 100
1166
1325
1356
3
12
37
1
19
7
CH2 twist
CH2 wag
740
1105
748 100
OH rock
1089
1222
1238
1359
3074
3620
93
0
18
21
1
CCl3 s str
CCl3 d str
CCl3 d str
C–O str
COH bend
O–H str
COH bend
CH2 scissor
CH2 sym str
CH2 asym str
O–H str
784
784
1113
1311
3604
778 100
1
2
3
469
976
079
1221
1388
786
1094
1266
3606
60
87
29
33
3
646
3652
28
3611
26
O–H str
a
UU
MP2r6-31G frequencies, scaled by 0.94.
and Ž9.0"0.8.=10y3 sy1 reported for these species
Fig. 2 shows the decay observed in the chamber
at Wuppertal. As seen from Fig. 2 the observed
decays followed first-order kinetics. Linear least-
in the Ford chamber w8x. CH ClOH decays faster in
2
the Wuppertal chamber than in the Ford chamber
squares analysis of the data in Fig. 2 gives first-order
while the reverse is true for CHCl OH. Such differ-
ences are consistent with a heterogeneous loss pro-
2
y3 y1
loss rates of Ž3.4"0.2.=10
s
for CH ClOH,
2
y3
y1
Ž5.5"0.3.=10
s
for CHCl OH, and Ž9.9"
cess as noted previously in comparisons of CH ClOH
2
2
y3
y1
0
.2. = 10
CH ClOH and CHCl OH in the Wuppertal chamber
s
for CCl OH. The decays of
decay rates in different chambers w8x. Heteroge-
neousrcatalytic processes are very sensitive to the
nature and history of the surface and can be very
discriminating, it is not unreasonable that there is a
3
2
2
y3
can be compared to decay rates of Ž1.6"0.1.=10
significant difference in the decay rates of CH ClOH
2
and CHCl OH in the different chambers.
2
The chamber at Wuppertal is fitted with a fan
which is used prior to irradiation to ensure complete
mixing of reactants. To provide additional informa-
tion concerning the relative importance of heteroge-
neous and homogeneous decomposition processes
for the chloromethanols additional experiments were
performed using the Wuppertal chamber. In these
experiments, the procedure was to introduce
CH OHrCl rN mixtures into the chamber with
3
2
2
the fan turned on for 2 min to ensure complete
mixing. The fan was then turned off, the mixture was
irradiated for 2–20 s to prepare the chloromethanols,
and an IR spectrum was recorded. The fan was then
turned on for 2 min and the FTIR spectrometer was
used to quantify the loss of chloromethanol. Use of
the fan to mix the contents of the chamber for 2 min
resulted in the complete Ž)97%. loss of each of the
Fig. 2. Decay of CH ClOH, CHCl OH, and CCl OH in the
chamber at Wuppertal when reaction mixtures were allowed to
stand in the dark Žfan switched off..
2
2
3

T.J. Wallington et al.rChemical Physics Letters 322 (2000) 97–102
101
UU
‡
Fig. 3. MP2r6-31G optimized structures of CCl OH Ža. wCCl OHx Žb., and HCl and CCl O Žc.. Distances in A˚ ngstroms and angles in
3
3
2
degrees.
chloromethanols. By comparison with Fig. 2, it can
be seen that the decay rates increased dramatically
when a fan was used to increase the circulation and
hence exposure of the chloromethanols to the reactor
walls. We conclude that heterogeneous decomposi-
tion of the chloromethanols on the chamber walls is
efficient and that the decay rates derived from Fig. 2
should be regarded as upper limits for the rates of
homogeneous decomposition.
CCSDŽT.rcc-pVTZ calculations predict the overall
y1
reaction enthalpy to be y12 kcal mol
at 298 K
ŽTable 2.. The enthalpy of HCl elimination for
CH ClOH and CHCl OH have previously been
2
2
y1
computed to be 7.8 and y2.5 kcal mol w8x, respec-
tively, reflecting the decreasing thermodynamic sta-
bility with increasing chlorination. The transition
state structure for HCl elimination from CCl OH is
3
qualitatively similar to that reported previously for
To compliment the experimental study, the first-
elimination of HF from CF OH w20x: the bond to the
3
order rate constant for HCl elimination from CCl OH
leaving halogen atom is substantially stretched and
the ClCO and HOC angles closed, bring the leaving
HCl close together. The large internal distortions in
the transition state suggest a high activation barrier
for homogeneous decomposition, which is borne out
by the energetic results. The activation energy is
3
was estimated using transition state theory with pa-
rameters determined from ab initio calculation. Fig.
UU
3
shows the MP2r6-31G optimized structures of
‡
CCl OH and wCCl OHx , the transition state for HCl
3
3
elimination, as well as the product HCl and CCl O.
2
UU
y1
Both the MP2r6-31G
and more sophisticated
calculated to be greater than 31 kcal mol , with the
Table 2
Calculated electronic, zero point, and thermal energies for the decomposition of CCl OH
3
U
CCl OH
wCCl OHx
CCl O
HCl
3
3
2
Absolute energies Žau.
UU
MP2r6-31G
y1492.477851
y1492.946804
0.023452
y1492.414376
y1492.890288
0.016860
y1032.276123
y1032.623344
0.010341
y460.215621
y460.337165
0.006883
CCSDŽT.rcc-pVTZ
a
ZPE
b
D H298
0.005412
0.003872
0.002360
Relative energies Žkcal moly 1
MP2qZPE
.
0.0
0.0
35.7
31.3
y12.6
y12.1
y12.5
y12.0
b
qD H
2
98
CCSDŽT.qZPE
b
qD H
2
98
a
UU
MP2r6-31G zero-point vibrational energy scaled by 0.9661.
b
UU
Thermal energy correction to 298 K, from MP2r6-31G structures and vibrational spectra.

1
02
T.J. Wallington et al.rChemical Physics Letters 322 (2000) 97–102
y1
CCSDŽT.rcc-pVTZ result several kcal mol below
References
the MP2 one. Combining the lower estimate for the
energy barrier with the MP2r6-31G structures and
vibrational spectra in a standard transition state the-
w1x M.J. Molina, F.S. Rowland, Nature 249 Ž1974. 810.
UU
w2x J.D. Farman, B.G. Gardiner, J.D. Shanklin, Nature 315 Ž1985.
2
07.
ory treatment w21x, we obtain a first-order rate con-
w3x T.J. Wallington, W.F. Schneider, D.R. Worsnop, O.J. Nielsen,
y10
J. Sehested, W. DeBruyn, J.A. Shorter, Environ. Sci. Tech.
stant for homogeneous decomposition of 2=10
y1
28 Ž1994. 320A.
w4x E.O. Edney, B.W. Gay Jr., D.J. Driscoll, J. Atmos. Chem. 12
s
, far less than the upper limit inferred from the
experiments above. Even allowing for a generous
estimate of possible overestimation of the barrier
height, and for contributions of quantum mechanical
tunneling, it is clear that the rate of homogeneous
Ž1991. 105.
5
w x
T.J. Wallington, M.D. Hurley, J.C. Ball, E.W. Kaiser, Envi-
ron. Sci. Tech. 26 Ž1992. 1318.
w6x E.C. Tuazon, R. Atkinson, J. Atmos. Chem. 17 Ž1993. 179.
w7x A.S. Hasson, I.W.M. Smith, J. Phys. Chem. A 103 Ž1999.
decomposition of CCl OH is slow, and that its life-
3
2
031.
time is quite long. The computational results are
consistent with the upper limits inferred from the
experiments above, and suggest that the lifetime of
w8x G.S. Tyndall, T.J. Wallington, M.D. Hurley, W.F. Schneider,
J. Phys. Chem. 97 1993 1576.
Ž
.
w9x J. Sehested, T.J. Wallington, Environ. Sci. Tech. 27 Ž1993.
1
46.
CCl OH is likely many days.
3
w10x I. Barnes, K.H. Becker, E.H. Fink, A. Reimer, F. Zabel, H.
Niki, Int. J. Chem. Kinet. 15 Ž1983. 631.
w11x T.J. Wallington, S.M. Japar, J. Atmos. Chem. 9 Ž1989. 399.
5
. Conclusion
12
w x
M.W. Schmidt, K.K. Baldridge, J.J. Boatz, S.T. Elbert, M.S.
Gordon, J.H. Jensen, S. Koseki, N. Matsunaga, K.A. Nguyen,
S.J. Su, T.L. Windus, M. Dupuis, J.A. Montgomery, J.
Comput. Chem. 14 Ž1993. 1347.
The experimental and computational results pre-
sented here provide a consistent picture of the stabil-
ity and IR spectra of CH ClOH, CHCl OH, and
CCl OH. All three compounds have lifetimes of at
least 100 s Žand probably much longer. with respect
to homogeneous decomposition in the gas phase, but
can decompose rapidly on surfaces.
w13x W.F. Schneider, B.I. Nance, T.J. Wallington, J. Am. Chem.
2
2
Ž
.
Soc. 117 1995 478.
3
w
14 A.P. Scott, L. Radom, J. Phys. Chem. 100 1996 16502.
x Ž .
w15x T.H. Dunning Jr., J. Chem. Phys. 90 Ž1989. 1007.
w16x J.F. Stanton, J. Gauss, J.D. Watts, W.J. Lauderdale, R.J.
Bartlett, ACES II, An Ab Initio Program System, Quantum
Theory Project, University of Florida.
w17x H. Kunttu, M. Dahlqvist, J. Murto, M. R a˜ s a˜ nen, J. Phys.
Chem. 92 Ž1988. 1495.
Acknowledgements
w18x M.C. Lin, J. Phys. Chem. 76 Ž1972. 811.
w19x P. Calza, C. Minero, E. Pelizzetti, Environ. Sci. Technol. 31
Ž1997. 2198.
We thank Willi Nelson ŽBergische Universit ¨a t
Wuppertal. for his help during these experiments.
T.J.W. acknowledges thanks the Alexander von
w20x W.F. Schneider, T.J. Wallington, R.E. Huie, J. Phys. Chem.
1
00 Ž1996. 6097.
w21x K.J. Laidler, Theories of Chemical Reaction Rates, Krieger,
Humboldt ŽAvH. Stiftung for an AvH fellowship.
New York, 1979.