Hydrolytic removal of the chlorinated products from the oxidative free-radical-induced degradation of chloroethylenes: Acid chlorides and chlorinated acetic acids

Article abstract of DOI:10.1039/b101687n

Progressive hydrolytic decomposition of acyl chlorides, among them the chlorinated acetyl chlorides, which are produced in the gas-phase oxidation of chlorinated ethylenes, permits the complete mineralization of organically bound chlorine to chloride anion. Hydrolysis rate constants (100% water) have been determined for the following acyl chlorides: acetyl (350 s^-1), chloroacetyl (5.5 s^-1), dichloroacetyl (300 s^-1), trichloroacetyl (>350 s^-1), and oxalyl dichloride (>350 s^-1). The chlorinated acetyl chlorides thereby give rise to the chloroacetates whose decomposition has also been studied and the kinetic parameters determined. Mono- and dichloroacetate anion undergo hydrolytic dechlorination (k_obs = k_o + k_OH- × [OH^-]; ClCH₂C(O)O^-: A_o 6.4 × 10¹⁵ s^-1, E_o 148 kJ mol^-1, A_OH- 1.6 × 10⁹ dm³ mol^-1 s^-1, E_OH- 86 kJ mol^-1. Cl₂CHC(O)O^-: A_o 3.2 × 10¹⁶ s^-1, E_o 156 kJ mol^-1, A_OH- 3.2 × 10¹⁰ dm³ mol^-1 s^-1, E_OH- 104 kJ mol^-1). Trichloroacetate anion decomposes by another mechanism, undergoing decarboxylation which is base-uncatalyzed: A_o 2.1 × 10¹⁷ s^-1, E_o 146 kJ mol^-1. Procedures on a pilot-plant scale are pointed out that allow the elimination of these compounds upon oxidation of the strip-gas produced when contaminated water is freed from chlorinated ethylenes by air-stripping.

Full text of DOI:10.1039/b101687n

View Article Online / Journal Homepage / Table of Contents for this issue
Hydrolytic removal of the chlorinated products from the oxidative
free-radical-induced degradation of chloroethylenes: acid chlorides
and chlorinated acetic acids
2
Lutz Prager,a Peter Dowideit,b Helmut Langguth,a Heinz-Peter Schuchmannb and
Clemens von Sonntagb
a Institut für Oberflächenmodifizierung, Permoserstrasse 15, D-04318 Leipzig, Germany
b Max-Planck-Institut für Strahlenchemie, Stiftstrasse 34-36, P.O. Box 101365,
D-45470 Mülheim an der Ruhr, Germany
Received (in Cambridge, UK) 20th February 2001, Accepted 28th March 2001
First published as an Advance Article on the web 25th April 2001
Progressive hydrolytic decomposition of acyl chlorides, among them the chlorinated acetyl chlorides, which are
produced in the gas-phase oxidation of chlorinated ethylenes, permits the complete mineralization of organically
bound chlorine to chloride anion. Hydrolysis rate constants (100% water) have been determined for the following
acyl chlorides: acetyl (350 sϪ1), chloroacetyl (5.5 sϪ1), dichloroacetyl (300 sϪ1), trichloroacetyl (>350 sϪ1), and oxalyl
dichloride (>350 sϪ1). The chlorinated acetyl chlorides thereby give rise to the chloroacetates whose decomposition
has also been studied and the kinetic parameters determined. Mono- and dichloroacetate anion undergo hydrolytic
Ϫ
Ϫ
9
3
dechlorination (kobs = ko ϩ kOH × [OH ]; ClCH2C(O)O : Ao 6.4 × 1015 sϪ1, Eo 148 kJ molϪ1, AOH 1.6 × 10 dm
Ϫ
Ϫ
molϪ1 sϪ1, EOH 86 kJ mol . Cl2CHC(O)O : Ao 3.2 × 1016 sϪ1, Eo 156 kJ molϪ1, AOH 3.2 × 1010 dm3 molϪ1 sϪ1, EOH
Ϫ1
Ϫ
Ϫ
Ϫ
Ϫ
104 kJ molϪ1). Trichloroacetate anion decomposes by another mechanism, undergoing decarboxylation which is
base-uncatalyzed: Ao 2.1 × 1017 sϪ1, Eo 146 kJ molϪ1. Procedures on a pilot-plant scale are pointed out that allow the
elimination of these compounds upon oxidation of the strip-gas produced when contaminated water is freed from
chlorinated ethylenes by air-stripping.
HC(O)Cl → CO ϩ Hϩ ϩ ClϪ
(1)
(2)
Introduction
Some of the most widespread environmental contaminants,
HC(O)Cl ϩ H2O → HC(O)OH ϩ HCl
especially of ground-water, are found among the chlorinated
ethylenes.1,2 Air-stripping has proved to be an efficient method
hydrolysis becomes competitive in strongly alkaline media
[reaction (3), k = 2.5 × 104 dm3 molϪ1 Ϫ1]. These are fast pro-
of ground-water remediation. The destruction of these com-
pounds in the strip-gas is effected by electron-beam treat-
ment3–5 or UV-photolysis.6–8 Under these conditions, part of
the decomposition products consists of acyl chlorides such as
mono-, di-, and trichloroacetyl chloride, phosgene, and formyl
chloride.3–5 These compounds can be removed from the air
stream by alkaline washing and hydrolysis. Although the strip-
gas contains water vapour, the hydrolysis must nevertheless be
conducted in the liquid phase since the reaction of acid chlor-
ides with water is slow in the gas phase.9 Estimation of the
rate of and optimization of the washing-out of these com-
pounds from the air stream leaving the reactor require data on
the mass transfer from the gas to the liquid phase. The rates of
transfer of the acid chlorides and of the subsequent hydrolytic
reactions are determining factors of this process and, therefore,
important aspects of this paper.
s
cesses from a technological point of view.
HC(O)Cl ϩ OHϪ → HC(O)OϪ ϩ Hϩ ϩ ClϪ (3)
The hydrolysis of phosgene is a two-step process. The rate-
determining step is the formation of carbonic acid mono-
12,13
chloride, slow [reaction (4), k ≈ 10 sϪ1
]
at neutral pH, but
fast in 1 M NaOH [reaction (5), k = 2.8 × 104 Ϫ1].13 The
s
subsequent decay of carbonic acid monochloride into carbon
dioxide and hydrogen chloride is fast [reaction (6); expected:
k ≥ 105 sϪ1, cf. refs. 10, 14–16].
C(O)Cl2 ϩ H2O → ClC(O)OH ϩ Hϩ ϩ ClϪ (4)
C(O)Cl2 ϩ OHϪ → ClC(O)OH ϩ ClϪ
ClC(O)OH → CO2 ϩ Hϩ ϩ ClϪ
(5)
(6)
The hydrolysis of the acid chlorides effects the removal of
these contaminants from the strip-gas but does not necessarily
achieve final detoxification. Chloroacetic acid chlorides are
hydrolyzed to toxic chloroacetates, which must be further
treated to achieve the transformation of the carbon-bound
chlorine to chloride.7
Recent publications have been concerned with the study of
the uptake of haloacetyl and carbonyl halides with the aim of
modelling cloud/atmosphere hydrolysis processes,17–21 whose
principles are very briefly gone into here. The net flux J of
gaseous molecules into the aqueous phase is quantitatively
characterized [expression (7)] by the mass uptake coefficient γ,
Dechlorination rate data concerning some of the acyl chlor-
ides of interest here exist in the literature. Formyl chloride in
aqueous solution mostly suffers non-hydrolytic HCl elimin-
ation [reaction (1), k = 104
s
Ϫ1].10,11 While near-neutrality
reaction (2) contributes only a few percent, OHϪ-catalyzed
J = nguγ/4
(7)
DOI: 10.1039/b101687n
J. Chem. Soc., Perkin Trans. 2, 2001, 1641–1647
1641
This journal is © The Royal Society of Chemistry 2001

View Article Online
the concentration ng in molecules per unit volume, and the
mean molecular velocity u. The factor of ¼ reflects the fact that
only half of the molecules approach the interface, and in a
more or less slanted direction. The quantity γ (cf. ref. 20)
increases with (i) the mass accommodation coefficient α, which
is defined as the fraction of sticky collisions relative to the total,
(ii) the solubility (which for these compounds decreases with
increasing ionic strength), and (iii) the rate of hydrolysis.
Details are discussed in refs. 17 and 20. A relatively low rate of
hydrolysis may lead to the saturation of the boundary layer
which impedes mass transfer. Such a situation is reflected, for
instance, in the data obtained for phosgene uptake,12 with
γ < 1.5 × 10Ϫ6 for water but, at 8 × 10Ϫ5, more than 40 times
larger for 1 M NaOH.17 In the case of trichloroacetyl chloride,
the mass uptake coefficient is about ten times larger than for
phosgene, as can be derived from the data given in refs. 21 and
22. Based on the knowledge of the quantities in expression (7)
and bearing in mind that the amount of mass transferred
depends on the contact time of the gas (which is inversely pro-
portional to the strip-gas flow rate), a scrubber can be designed
with the dimensions such that it achieves the desired scrubbing
trichloroacetate decomposes by a unimolecular pathway
whereby chloroform is released [reactions (12) and (13)].31 In
the latter case, the mineralization of the organic chlorine
requires the subsequent hydrolysis of the chloroform [reactions
(14), (15) and (1)].
ClCH2C(O)OϪ ϩ OHϪ → HOCH2C(O)OϪ ϩ ClϪ (9)
Cl2CHC(O)OϪ ϩ OHϪ → ClCH(OH)C(O)OϪ ϩ ClϪ (10)
ClCH(OH)C(O)OϪ → HC(O)C(O)OϪ ϩ Hϩ ϩ ClϪ (11)
CCl3C(O)OϪ → CCl3Ϫ ϩ CO2
CCl3Ϫ ϩ H2O → HCCl3 ϩ OHϪ
CCl3H ϩ OHϪ → HOCHCl2 ϩ ClϪ
HOCHCl2 → CH(O)Cl ϩ HCl
(12)
(13)
(14)
(15)
The alkali-catalyzed rate of hydrolysis of chloroform32 is
faster than that of monochloroacetate and dichloroacetate
under the same conditions (at lesser alkalinities it may be slower
than the decomposition of trichloroacetate [reaction (12)]
which is not alkali-catalyzed). Thus, if the decomposition of
the chloroacetates is carried out in strongly alkaline solution
chloroform will not accumulate. Dichloromethanol promptly
eliminates HCl [reaction (15)]10 producing formyl chloride,
which decays very quickly under these conditions [reactions (1)
and (2)].10,11
ratio cgas,final/cgas,initial
.
In the present paper, rate determinations are reported regard-
ing the hydrolysis of acetyl chloride and halogenated acetyl
chlorides [reaction (8)] in aqueous mixed-solvent systems. In
ClnCH3 Ϫ nC(O)Cl ϩ H2O →
ClnCH3 Ϫ nC(O)OϪ ϩ 2 Hϩ ϩ ClϪ (n = 0, 1, 2, 3) (8)
100% water, the hydrolysis of aliphatic acyl chlorides is usually
so fast that its rate cannot be followed by slow conventional
techniques. To reduce the rate, but also to elucidate the
hydrolysis mechanism with respect to the order of the kinetics
as a function of the water concentration, organic cosolvents
and low temperatures have often been used,23–27 but such data
do not readily allow an extrapolation to the neat-water situ-
ation at room temperature. Thus, fast kinetics techniques must
be applied. There are some past examples of this. The rate of
the hydrolysis of acetyl chloride in water-rich solvent systems
has been studied kinetically by a stopped-flow method.28 The
hydrolysis of phosgene has been studied by shooting a water jet
through a phosgene atmosphere; from the HCl taken up, the
rate of phosgene hydrolysis was estimated by applying non-
equilibrium absorption theory.12 Phosgene has been generated
in situ using the pulse radiolysis technique, and its hydrolysis
followed by fast conductometry.13 This has allowed the study of
a wider temperature range and has yielded values for the rate
constants in a straightforward way as no additional assump-
tions are needed in this case since the system is monophasic.
Succinic dichloride15 and formyl chloride10 can be similarly
generated in situ. A study of the hydrolysis of trichloroacetyl
chloride reports rate constants in the order of between 102
It is thus seen that complete mineralization of organic chlor-
ine in these systems is feasible. In the present work, further
kinetic studies of the hydrolysis of these compounds are under-
taken with a view to developing an optimized procedure that
operates on a pilot-plant scale.29
Experimental
Stopped-flow experiments
All acyl chlorides were of the highest purity commercially
available and used without further purification. 1,4-Dioxane,
the organic component of the mixed-solvent systems, was
dried over sodium wire and redistilled. The Milli-Q-filtered
(Millipore) water had a background resistivity of 18 MΩ cmϪ1
.
In the course of our studies on the reaction of ozone with
chlorinated olefins,33 a fast small-volume conductometric detec-
tion system was constructed from glass tubing (the cell volume,
containing the two mutually opposite glassy-carbon electrodes
of 1 mm diameter, was only 30 µl),11 and fitted to a stopped-
flow set-up (Biologic SFM-3), which allowed the measurement
of conductance changes with a time resolution of 2 ms, i.e.
within times very close to the mixing time, corresponding to
rate constants of up to 350 sϪ1. This arrangement is about five
times faster than that of a similar conductometric set-up.34 The
conductance change was measured with a Wheatstone-bridge
arrangement adapted from a pulse radiolysis detection system35
against a reference cell filled with a fully reacted sample from a
preceding run (for details see ref. 11). With the help of this
set-up the rate constants for a series of acyl chlorides were
determined in 1,4-dioxane–water mixtures, where the acyl
chlorides were dissolved in dry 1,4-dioxane and mixed with
aqueous dioxane of the required water content. It was possible
to extend the experiments to such low 1,4-dioxane
concentrations that a meaningful extrapolation to the neat-
water situation was considered feasible.
and 103 Ϫ1, using the stopped-flow and the droplet-train
s
techniques.22 For comparison, our own result is that this rate
constant is considerably in excess of 350 sϪ1 (see below).
From the foregoing discussion it can be assumed that
these acid chlorides should, in principle, be easily hydrolyzed,
especially by the use of alkaline washing liquids, once they have
penetrated into the aqueous phase. The rate of acid-chloride
transfer from the gas to the aqueous phase must be kept in
mind; a pilot-plant packed-column absorber has been used29
to determine the overall acid-chloride removal efficiency (see
below).
Harsher conditions must be employed for the transformation
of the chloroacetates into innocuous end products. The alkaline
hydrolysis of monochloroacetate into glycolate30 and of
dichloroacetate into glyoxylate31 [reactions (9) and (10), (11)]
shows second-order kinetics {the first step being rate-
determining while the hydrolysis of the chlorohydroxyacetate
[reaction (11)], a geminal chlorohydrin, is fast}. In contrast,
The fast progress of the hydrolysis of acetyl chloride is shown
in Fig. 1. The acidification of the solutions in the course of the
hydrolysis does not interfere with the kinetics since acid
catalysis is practically absent in this class of compounds (see
below).
1642
J. Chem. Soc., Perkin Trans. 2, 2001, 1641–1647

View Article Online
Fig. 1 Kinetic trace (conductivity vs. time) of the hydrolysis of
acetyl chloride ([H2O] = 35 mol dmϪ3 in 1,4-dioxane). The initial period
of the mixing event with its concomitant noisy signal has been omitted.
Fig. 2 Rate of hydrolysis of acetyl chloride in 1,4-dioxane–water
mixtures. Data from ref. 28 (*) are also included.
Hydrolytic absorber for the acyl chlorides
Simulated oxidized strip-gas, i.e. air loaded with the acid chlor-
ide to be studied [the acid chlorides were produced from
trichloro-, tetrachloro-, or (E)-1,2-dichloroethylene by UV (222
nm) photooxidation] at a flow rate in the order of 102 m3 hϪ1
enters a pilot-plant-scale scrubber with one or two (optional
in series) packed columns (supplier: Dr Günther, Engineer-
ing, Wittenberg, Germany), packed volume 1.25 × 0.33 m,
filled with PallTM rings, specific surface 220 m2 mϪ3 (supplier
Vereinigte
Füllkörper-Fabriken,
Ransbach-Baumbach,
Germany). 0.5–1.5 M NaOH (0.35 m3, circulated through the
absorber at a flow rate of 0.5–2.3 m3 hϪ1) is sprayed in at the
top. The set-up has previously been described in detail.29 These
experiments take into account the fact that the rate of mass
transfer from the gas to the aqueous phase is an important
parameter in the removal of the acid chlorides from the
strip-gas. The disappearance of the acid chlorides was followed
on-line by FTIR (Fourier-transform infrared) spectroscopy
(Bruker Vector 22; windows NaCl, optical path length 0.2 m,
detection limit for the acid chlorides about 1 ppm).29
Fig. 3 Rate of hydrolysis of chloroacetyl chloride in 1,4-dioxane–
water mixtures.
dissolved in dry 1,4-dioxane, which is fully miscible with water.
The rate of hydrolysis was observed to increase with increasing
water concentration, but not in a linear fashion over the full
concentration range studied. Examples are shown in Figs. 2 and
3. In the case of acetyl chloride (Fig. 2) three domains may be
distinguished: in the dioxane-rich region the rate of hydrolysis
remains low up to about 22 mol dmϪ3 water after which the rate
increases steeply; between about 25 and 50 mol dmϪ3 water the
rate increases linearly with the water content. It is expected that
this linear part of the rate vs. [H2O] relationship may be
extrapolated to 100% water, i.e. that no further discontinuity
occurs in the range between 90 and 100% water content. The
uneven trend of these results is in contradiction to that reported
in the literature28 for the same solvent system where the rate
constant rises in an almost exponential manner with the water
concentration (symbols * in Fig. 2). If one accepts the results
obtained in the present work, the extrapolation leads to a
hydrolysis rate constant of 350 sϪ1 for acetyl chloride in 100%
water. This extrapolation seems justified since the gap between
the last point measured (90% water) and the dioxane-free sys-
tem is small, and one might expect a continuation of the linear
change of the variable in question with the molar fraction of
water (provided that the interaction of a 1,4-dioxane molecule
with the acetyl chloride molecule is relatively weak, which is
probably the case).
Alkaline hydrolysis of chloroacetates
Batch-type, bench-scale experiments were carried out to
determine the degradation rate constants of mono-, di-, and
trichloroacetates as a function of NaOH concentration (up to
1 M) and temperature (range from 80 to 120 ЊC). The samples
of chloroacetate concentrations in the millimolar range were
hydrolyzed in closed glass or PTFE vessels capable of sustain-
ing the pressure that builds up when the aqueous solution is
heated to above 100 ЊC, or of resisting attack by highly alkaline
solutions.29 The samples were kept in a heated oil bath for the
desired time, chilled after the run, diluted with distilled water,
and analyzed for the remaining chloroacetate and for the ionic
products by ion chromatography (Dionex, column AS12A 4
mm, eluent 50 mM sodium hydroxide).
Experiments to determine the conditions needed to achieve
degradation of the chloroacetates so as to conform to the
maximum statutory limit for disposal in the municipal sewage
system were carried out in a pilot-plant-scale flow-reactor,
which was run at temperatures of up to 210 ЊC. Details regard-
ing its construction and operation have already been reported.29
Extrapolation of the results reported by Hudson and Moss
in ref. 28 (their28 Fig. 3 and taking the Grunwald–Winstein
parameter36 Y(100% H2O) equal to 3.6,36 together with
Results and discussion
Acyl-chloride hydrolysis rate constants
their28 Table 3) leads to a rate constant of 1200 sϪ1
.
The rate of hydrolysis of some aliphatic acyl chlorides has been
determined by the stopped-flow technique with conductometric
detection. The upper detection limit of this setup is 350 sϪ1
(equivalent to a mixing-time of 2 ms). The acyl chlorides were
While the rate of hydrolysis of chloroacetyl chloride in the
higher water-concentration range is much lower, the overall
picture at the higher molar fractions of water is qualitatively the
same (Fig. 3). Extrapolation leads to a hydrolysis rate constant
J. Chem. Soc., Perkin Trans. 2, 2001, 1641–1647
1643

View Article Online
Table 1 Rate of hydrolysis of some aliphatic acyl chlorides at 20 ЊC
(100% water)
Acyl chloride
Phosgene
khydrolysis
Ref.
6 sϪ1
12
13
13
9 sϪ1
[OHϪ] × 2.8 ×
104 dm3 molϪ1 sϪ1
a
Formyl chloride
(104 sϪ1
)
10
10, 37
10
~400 sϪ1
[OHϪ] × 2.5 ×
104 dm3 molϪ1 sϪ1
350 sϪ1
been inferred by IR spectroscopy in dioxane solutions of low
water content.26
Acetyl chloride
This work
This work
This work
This work
22
This work
15
Chloroacetyl chloride
Dichloroacetyl chloride
Trichloroacetyl chloride
5.5 sϪ1
Hydration [reaction (18)] is perhaps a concerted process that
involves more than one water molecule, as exemplified in reac-
tion (20) whose transition-state might include two molecules of
water.
300 sϪ1
>350 sϪ1
150–500 sϪ1
>350 sϪ1
6 sϪ1
Oxalyl dichloride
Succinyl dichloride
a Mainly decarbonylation [reaction (1)].
(20)
of 5.5 sϪ1 for chloroacetyl chloride in 100% water. Rate
constants for the hydrolysis of these and several other aliphatic
acyl chlorides also measured in the present study are compiled
in Table 1, together with some additional data from the
literature.
Reactions that occur in aqueous solutions are often
facilitated by the co-operation of a cluster of water molecules
even though they might not be hydrolytic. A case in point is the
very rapid decomposition of formyl chloride which occurs in
water many orders of magnitude faster than in the gas phase.10
The termination of formaldehyde radical anion appears to be
another example of the involvement of more than one water
molecule in the reaction.50
In the series of aliphatic acyl chlorides its first member,
formyl chloride, decomposes relatively rapidly into CO and
HCl [reaction (1)], such that the competing hydrolysis into
formic acid and HCl [reaction (2)] is difficult to assess with
accuracy,10 but from our earlier10 and more recent37 data a
value close to 400 sϪ1 seems to be reasonable. While acetyl
chloride in 100% water hydrolyzes with a rate constant of 350
Assuming that acetyl chloride decomposes according to an
SN1 mechanism in media of sufficiently high ionizing power,
electron-drawing substituents which disfavour the release of a
chloride ion are thus expected to hamper the SN1 reaction. In
agreement with this expectation the hydrolysis rate of chloro-
acetyl chloride is about two orders of magnitude slower
than that of the unsubstituted compound. However, with the
dichloro compound, the rate is found to be greater again which
might indicate a change of mechanism, and further substitution
with chlorine increases the rate of hydrolysis even more (Table
1). It might be argued that this can be attributed to an increase,
from the mono- to the trichloroacetyl chloride, in the rate
of hydrate formation. An analogy suggests itself: one recalls
that while acetaldehyde and its hydrate are present in water at
about equal concentrations, its trichloro derivative is almost
exclusively hydrated (“chloral hydrate”). The irregular trend of
the rate constants as regards the 100%-aqueous solutions is in
contrast with the systems of low water content where the
rate rises regularly with the degree of chlorine substitution
(cf. Figs. 2 and 3, and ref. 25).
s
Ϫ1, substitution with one chlorine reduces the rate to 5.5 sϪ1
;
further chlorine substitution however re-enhances the rate, to
300 sϪ1. The rate of hydrolysis of trichloroacetyl chloride is so
fast that it exceeds the detection limit (350 sϪ1) of the present
set-up. Similarly, a study of the hydrolysis of trichloroacetyl
chloride under atmospheric/cloud conditions reports rate con-
stants in the order of between k = 150 sϪ1 (pH 4.8–5.8) using the
stopped-flow technique and k = 500 sϪ1 (below pH 3.4) with the
droplet-train technique.22 The discontinuity observed on going
from acetyl chloride to monochloroacetyl chloride clearly
indicates a change in the mechanism of hydrolysis.
Mechanistic aspects
The present work is focused on hydrolysis rates in purely aque-
ous systems; the dioxane–water system used here provides the
basis for extrapolation to the 100%-water situation where the
rate may be too fast to measure in our set-up. Compared with
purely aqueous solutions, mixed-solvent systems usually have
the additional complexity of hydrophobic binding of the
organic component with the substrate molecule. This implies,
for instance, that the dependence of the rate of hydrolysis on
the water content reflects variations of accessibility of the water
molecules to the substrate molecule as well as a possible shift
(with higher water content) toward the SN1 pathway, or the
consequences of molecularity ([H2O]n).
For the non-catalyzed hydrolysis of acyl chlorides, mainly
two mechanisms have been considered, i.e. SN1 [reactions (16)
and (17)], and SN2, which may proceed via a hydrate [reactions
(18) and (19)],38–41 while solvent–acyl chloride charge-transfer
complexes have often been considered as well. Arguments in
favour of the relative importance of these different pathways
have been reviewed.42–46 The rate-limiting steps will be reactions
(16) and (18), since the lifetimes of carbocations in water, even
when stabilized by electron-donating substituents, are at the
most a few µs,47–49 and the decomposition of geminal chloro-
hydrins in water is also very fast (in the order of 106 sϪ1).14–16
The existence of acetyl chloride hydrate as an intermediate has
Regarding the catalysis of hydrolysis reactions in acyl
halides, the nucleophilic pathway (OHϪ) is more effective
than the electrophilic one (e.g., Hϩ). In the case of acetyl chlor-
ide, the catalytic effect of acid is apparently nil,51 while it
is observable in trichloroacetyl chloride at pH below about 5.22
There is also a catalytic effect in the case of acyl fluorides where
the proton appears to attach to the fluorine halide which is then
removed as hydrogen fluoride.42,45 OHϪ reacts by addition at the
carbonyl carbon, followed by release of the halide ion.
The rate constants of the OHϪ-induced hydrolysis of phos-
gene13 and formyl chloride10 and the OHϪ-induced hydration
of acetaldehyde52 are in the order of 104–105 dm3 molϪ1 sϪ1
,
i.e. slow compared to the deprotonation at a heteroatom (e.g.
at O–H) which are close to diffusion controlled.53 Assuming a
maximum rate of OHϪ-addition of 105 dm3 molϪ1 sϪ1, this sets
the limit for a contribution of OHϪ at pH 7 at 10Ϫ2 Ϫ1. All of
s
the rates measured in the present study are well above this value,
and even in the slowest ones the kinetics do not change as the
pH drops in the course of the reaction. As mentioned above,
1644
J. Chem. Soc., Perkin Trans. 2, 2001, 1641–1647

View Article Online
Fig. 4 Absorption of phosgene from an air stream containing
products from the photooxidation of chlorinated ethylenes as a func-
tion of absorber-column contact time and absorbent pH. [C(O)Cl2]o
~50 ppm. pH 13.26 (᭹), 13.61 (᭺), 13.80 (ꢀ), 13.90 (᭝).
Fig. 5 Kinetics of the hydrolysis of dichloroacetate at 80 ЊC (᭹),
100 ЊC (ᮀ) and 120 ЊC (ꢀ) as a function of the OHϪ concentration.
any Hϩ-catalyzed hydration as a prelude to hydrolysis can be
disregarded. Thus, it is justified to discuss the observed rates on
the basis of the reactions (16)–(19).
Pilot-plant-scale acyl chloride hydrolysis
Trichloroethylene, tetrachloroethylene, and (E)-dichloro-
ethylene were photooxidized and the concentrations of the
products dichloroacetyl chloride, trichloroacetyl chloride,
phosgene, and formyl chloride, as well as carbon dioxide and
carbon monoxide, obtained at different alkalinities and gas flow
rates, were determined by FTIR spectroscopy in front, between,
and behind the two absorber columns of the set-up used.29 The
results for phosgene are shown in Fig. 4.
The logarithm of the ratio of exit to entry concentrations is
seen to decrease linearly with the contact time (i.e. a depend-
ence of the time constant of phosgene removal on its initial
concentration was not observed). It is pH-sensitive in the
(strongly alkaline) pH range investigated (Fig. 4). Under the
conditions that apply in the experiments referred to in Fig. 4,
the absorption of phosgene is incomplete. Here, the criterion
for “completeness” is a cgas,final of 0.2 ppm, i.e. about one fifth
of the statutory (in Germany) emission limit of 1 mg mϪ3. In
contrast, for the chlorinated acetyl chlorides whose mass
uptake coefficients γ are larger by about an order of magnitude
(see above), no column overflow was observed (detection limit,
cgas,final/cgas,initial < 0.02). This is in agreement with rough calcu-
lations based on the mass uptake coefficients γ (see above) and
data characterizing the reaction conditions in the scrubber such
as the gas–liquid interfacial area, residence time, and gas flow
rate.
Fig. 6 Arrhenius plot of the degradation of trichloroacetate (᭺,
water; ᭞, water, vessel silanised; ᭿, 5 M NaOH; ꢀ, 1.5 M NaOH, vessel
silanised; ᭜, ref. 31. Inset: kinetics of the degradation of trichloro-
acetate at 100 ЊC; ln kobs as a function of time at varying OHϪ concen-
trations (᭹, 5 × 10Ϫ3 mol dmϪ3 NaOH, ᭿, 1 × 10Ϫ2 mol dmϪ3 NaOH;
᭞, 0.1 mol dmϪ3 NaOH; ᭜, 0.5 × 10Ϫ3 mol dmϪ3 NaOH).
products chloride ion and glycolic acid give an essentially com-
plete material balance.29
The degradation rate constants of dichloroacetate and
trichloroacetate were obtained in a like manner (for data see ref.
29). In the case of dichloroacetate as in the monochloro com-
pound, the rate constant consists of two terms, with ln ko =
38.0 Ϫ 18800/T (Ea 156 kJ molϪ1) and ln kOH = 24.2 Ϫ 12500/T
Ϫ
(Ea 103.7 kJ molϪ1, in good agreement with the literature31).
Wall effects appear to play some role in ko.29 The degradation
essentially furnishes chloride and glyoxylic acid. A decarb-
oxylation reaction analogous to reaction (12) is unimportant
here (below 1%). The presence of glycolic and oxalic acids is
ascribed to the Cannizzaro reaction of glyoxylic acid under
these strongly alkaline conditions.
Thus, in technical applications, phosgene absorption pro-
vides the limiting factor in the design of a gas scrubber.
Alkaline degradation of chloroacetates
In contrast to the mono- and dichloroacetates, trichloro-
acetate decays by decarboxylation, owing to the strongly
electron-drawing trichloromethyl group, essentially uncatalyzed
by OHϪ. An Arrhenius plot is presented in Fig. 6 which shows
that the kinetics is not influenced by alkali or the nature of the
vessel walls.
The kinetics is described by ln ko = 39.9 Ϫ 17400/T (Ea 144.6
kJ molϪ1, in good agreement with the literature31). While the
presence of strong base is irrelevant for the degradation of the
trichloroacetate, it ensures the subsequent decomposition of
the chloroform resulting from the decarboxylation reaction
[cf. reactions (12) and (13)] according to an SN2 mechanism
[cf. reactions (14) and (15)].
The complete hydrolytic dechlorination of chloroacetic
acids, the products of the hydrolysis of the chlorinated acetyl
chlorides, requires harsher conditions (temperatures beyond
the boiling point of water up to 200 ЊC) than the hydrolysis
of the corresponding acid chlorides.7 Kinetic studies have been
done on monochloroacetate (SN2 kinetics),30 dichloroacetate
(SN2 kinetics),30,31 and trichloroacetate (decarboxylation).31
Fig. 5 shows, by way of example, some kinetic results of the
batch hydrolysis experiments of dichloroacetate.
In the cases of the mono- and dichloroacetates the rate
Ϫ
Ϫ
constant is made up of two terms, i.e. kobs = ko ϩ kOH × [OH ].
For monochloroacetate, the temperature dependence of the
intercept gives ln ko = 36.4 Ϫ 17800/T, corresponding to an
activation energy of 148 J molϪ1 (cf. Table 2; diagram not
shown). The Arrhenius plot of the second-order rate constant
Some comments can be made regarding the quantities in
Table 2, columns 2 and 5. The frequency factors in column 2, in
particular the bottom one, appear to be quite large. High fre-
quency factors may be associated with reactions in which more
kOH gives ln kOH = 21.2 Ϫ 10300/T (Ea 85.7 kJ molϪ1). The
Ϫ
Ϫ
J. Chem. Soc., Perkin Trans. 2, 2001, 1641–1647
1645

View Article Online
Table 2 Activation parameters of the dechlorination/degradation of chlorinated acetates
H2O-induced
H2O-induced
OHϪ-induced
OHϪ-induced
Chlorinated acetate
A/sϪ1
Ea/kJ molϪ1
A/dm3 molϪ1 sϪ1
Ea/kJ molϪ1
Chloroacetate
Dichloroacetate
Trichloroacetate
6.4 × 1015
3.2 × 1016
2.1 × 1017 a
148
156
146a
1.6 × 109
3.2 × 1010
—
86
104
—
a First-order but not H2O-induced.
in 1 M NaOH at 150 ЊC, or in water at 180 ЊC, i.e. there is a
trade-off between the operation of the process at relatively low
overpressures but in a relatively corrosive reaction medium, and
vice versa.
Acknowledgements
Partial financial support by the German Federal Ministry of
Education and Research (Project No. 02WT9656/57) and the
Stadtwerke Düsseldorf AG (Project No. 230/45083468) is
gratefully acknowledged.
References
1 R. Koppmann, F. J. Johnen, C. Plass-Dülmer and J. Rudolph,
J. Geophys. Res., 1993, 98, 20517.
2 A. McCulloch and P. M. Midgley, Atmos. Environ., 1996, 40, 601.
3 L. Prager, R. Mehnert, A. Sobottka, H. Langguth, W. Baumann,
W. M ätzing, H.-R. Paur, J. Schubert, R. Rashid, K. M. Taba,
H.-P. Schuchmann and C. von Sonntag, J. Adv. Oxid. Technol., 1998,
3, 87.
Fig. 7 Degree of degradation cmeasured/cmax of dichloroacetate (᭹)
without alkali; liberated chloride (᭿); sum (᭝) of 2 × dichloroacetate
and liberated chloride; contact time in the hydrothermal flow reactor,
480 s. Inset: trichloroacetate (᭹), liberated chloride (᭿), sum (᭝) of
3 × trichloroacetate and liberated chloride; contact time, 450 s.
4 T. Mill, M. G. Su, C. C. D. Yao, S. M. Matthews and F. T. S. Wang,
Radiat. Phys. Chem., 1997, 50, 283.
5 T. Hakoda, S. Hashimoto, J. Fujiyama and A. Mizuno, J. Phys.
Chem. A, 2000, 104, 59.
6 E. Sanhueza, I. C. Hisatsune and J. Heicklen, Chem. Rev., 1976, 76,
801.
than one bond is ruptured simultaneously, as well as certain
dissociation reactions where a bond is broken that links two
groups of atoms.54 The hydrolytic removal of chloride in the
mono- and dichloroacetates can be viewed as an example of the
first class, e.g. reaction (21), with a C–Cl and an O–H bond
7 W. R. Haag, M. D. Johnson and R. Scofield, Environ. Sci. Technol.,
1996, 30, 414.
8 L. Prager and E. Hartmann, J. Photochem. Photobiol., A, 2001, 138,
177.
ClCH2C(O)OϪ ϩ H2O →
HOCH2C(O)OϪ ϩ ClϪ ϩ Hϩ (21)
9 R. Butler and A. Snelson, J. Air Pollut. Control Assoc., 1979, 29,
being cleaved as a water molecule attaches to the chlorine-
bearing carbon atom. On the other hand, the decomposition of
trichloroacetate seems akin, for instance, to the homolysis of
833.
10 P. Dowideit, R. Mertens and C. von Sonntag, J. Am. Chem. Soc.,
1996, 118, 11288.
11 P. Dowideit, Doctoral thesis, Ruhr-Universität Bochum, 1996.
12 W. H. Manogue and R. L. Pigford, AiChE J., 1960, 6, 494.
13 R. Mertens, C. von Sonntag, J. Lind and G. Merényi, Angew. Chem.,
Int. Ed. Engl., 1994, 33, 1259.
14 R. Köster and K.-D. Asmus, Z. Naturforsch., B. Anorg. Chem. Org.
Chem., 1971, 26, 1108.
15 R. Mertens and C. von Sonntag, J. Chem. Soc., Perkin Trans. 2,
1994, 2181.
16 N. Latif, P. O’Neill, D. Schulte-Frohlinde and S. Steenken, Ber.
Bunsen-Ges. Phys. Chem., 1978, 82, 468.
17 W. J. De Bruin, S. X. Duan, X. Q. Shi, P. Davidovits, D. R. Worsnop,
M. S. Zahniser and C. E. Kolb, Geophys. Res. Lett., 1992, 19, 1939.
18 P. Davidovits, J. H. Hu, D. R. Worsnop, M. S. Zahniser and
C. E. Kolb, Faraday Discuss., 1995, 100, 65.
alkanes [e.g., C(CH3)4 → CH3 ϩ C(CH3)3 has A ~ 1017 sϪ1].54
The activation energy of the OHϪ-catalyzed dechlorination is
seen to increase on going from the mono- to the dichloroacetate
(Table 2, column 5). This is in agreement with the idea that
as the partial charge on the chloromethyl group increases, the
barrier for the entry of OHϪ also increases, so that finally in the
case of the trichloroacetate the hydrolytic pathway is practically
blocked, relative to decarboxylation. It is, moreover, reasonable
to assume that OHϪ attack is at the chloromethyl group not at
the carbonyl group, which in fact has lost its usual character-
istics owing to carboxylate mesomerism.
ؒ
ؒ
19 G. M. Nathanson, P. Davidovits, D. R. Worsnop and C. E. Kolb,
J. Phys. Chem., 1996, 100, 13007.
20 C. E. Kolb, J. T. Jayne, D. R. Worsnop and P. Davidovits, Pure Appl.
Chem., 1997, 69, 959.
21 W. J. De Bruyn, J. A. Shorter, P. Davidovits, D. R. Worsnop,
M. S. Zahniser and C. E. Kolb, Environ. Sci. Technol., 1995, 29,
1179.
22 C. George, J. Lagrange, P. Lagrange, P. Mirabel, C. Pallares and
J. L. Ponche, J. Geophys. Res., 1998, 99, 1255.
23 T. W. Bentley, G. Llewellin and J. A. McAlister, J. Org. Chem., 1996,
61, 7927.
24 V. Gold and J. Hilton, J. Chem. Soc., 1955, 3303.
25 I. Ugi and F. Beck, Chem. Ber., 1960, 94, 1839.
26 V. A. Terentev and V. V. Varfolomeeva, Russ. J. Gen. Chem., 1996,
66, 1954.
27 E. J. Cairns and J. M. Prausnitz, J. Chem. Phys., 1960, 32, 169.
28 R. F. Hudson and G. E. Moss, J. Chem. Soc., 1962, 5157.
Hydrothermal flow reactor
The results of some systematic decomposition studies with a
hydrothermal flow reactor29 (in which the disappearance of the
reactant and the appearance of products were followed as func-
tions of temperature and alkalinity) conform to expectation
based on the foregoing rate constants. This technique has
been applied to chloroacetates before.7 Practically complete
dechlorination (a typical example is shown in Fig. 7) of these
compounds is achievable within a time span in the order of
102 s at temperatures well below 200 ЊC.
In 1 M NaOH as the medium, the half-way point of the
dichloroacetate degradation is shifted to lower temperatures by
about 22 ЊC, i.e. to below 130 ЊC (data not shown). Thus, com-
plete dechlorination of dichloroacetate is achieved after 8 min
1646
J. Chem. Soc., Perkin Trans. 2, 2001, 1641–1647

View Article Online
29 L. Prager, Doctoral thesis, Leipzig University, Leipzig, Germany,
2000.
43 A. Kivinen, Acta Chem. Scand., 1965, 19, 845.
44 H. Minato, Bull. Chem. Soc. Jpn., 1964, 37, 316.
45 A. Kivinen, in The Chemistry of Acyl Halides, ed. S. Patai, Wiley
Interscience, London, 1972, p. 177.
30 H. Dautzenberg and B. Philipp, Z. Phys. Chem. (Leipzig), 1979, 260,
289.
31 H. Dautzenberg and B. Philipp, Z. Phys. Chem. (Leipzig), 1979, 260,
298.
32 P. M. Jeffers, L. M. Ward, L. M. Woytowitch and N. L. Wolfe,
Environ. Sci. Technol., 1989, 23, 965.
33 P. Dowideit and C. von Sonntag, Environ. Sci. Technol., 1998, 32,
1112.
34 H. Strehlow, Rapid Reactions in Solution, VCH, Weinheim, 1992.
35 E. Bothe and E. Janata, Radiat. Phys. Chem., 1994, 44, 455.
36 E. Grunwald and S. Winstein, J. Am. Chem. Soc., 1948, 70, 846.
37 A. Leitzke, R. Flyunt and C. von Sonntag, in preparation.
38 D. A. Brown and R. F. Hudson, J. Chem. Soc., 1953, 3352.
39 V. Gold, J. Hilton and E. G. Jefferson, J. Chem. Soc., 1954, 2756.
40 V. Gold and J. Hilton, J. Chem. Soc., 1955, 3303.
41 E. W. Crunden and R. F. Hudson, J. Chem. Soc., 1956, 501.
42 R. J. E. Talbot, in Ester Formation and Hydrolysis and Related
Reactions, ed. C. H. Bamford and C. F. H. Tipper, Elsevier,
Amsterdam, 1972, p. 209.
46 D. P. N. Satchell and R. S. Satchell, in Supplement B. The Chemistry
of Acid Derivatives, Vol. 2, ed. S. Patai, Wiley, New York, 1992,
p. 747.
47 S. Steenken, J. Buschek and R. A. McClelland, J. Am. Chem. Soc.,
1986, 108, 2808.
48 S. Steenken and R. A. McClelland, J. Am. Chem. Soc., 1989, 111,
4967.
49 D. J. Deeble, M. N. Schuchmann, S. Steenken and C. von Sonntag,
J. Phys. Chem., 1990, 94, 8186.
50 W.-F. Wang, M. N. Schuchmann, V. Bachler, H.-P. Schuchmann
and C. von Sonntag, J. Phys. Chem., 1996, 100, 15843.
51 V. Gold and J. Hilton, J. Chem. Soc., 1955, 838.
52 J. L. Kurz and J. I. Coburn, J. Am. Chem. Soc., 1967, 89,
3528.
53 M. Eigen, W. Kruse, G. Maass and L. De Maeyer, Prog. React.
Kinet., 1964, 2, 285.
54 C. Steel and K. J. Laidler, J. Chem. Phys., 1961, 34, 1827.
J. Chem. Soc., Perkin Trans. 2, 2001, 1641–1647
1647