View Article Online / Journal Homepage / Table of Contents for this issue

Phys. Org. 1727

Reactions of Keten. Part I. Kinetics of the Gas-phase Reaction with

Acetic Acid
By P. G. Blake and H. H. Davies, Chemistry Department, University College, Cardiff CF1 1XL
The addition of keten to acetic acid to give acetic anhydride is a homogeneous bimolecular gas-phase reaction
between 106 and 21 5 "C. The reaction is first-order in keten and in acid monomer and k = 105.e1exp (-46,260/-
R T ) I mol-l s-l.
Published on 01 January 1971. Downloaded by University of Kentucky on 13/01/2016 22:18:30.

DESPITE the industrial importance of the conversion of and B = (a - b)/a, a, b, and x having the usual mean-
acetic acid to acetic anhydride by keten no kinetic ings.a
studies appear to have been reported. Satchell and co- MateriaZs.-Acetic acid of better than 99.8% purity was
workers1 have studied the kinetics of addition of degassed before use. Keten was prepared by pyrolysing
acetic anhydride in a flow system at 400 "C, followed by
dimethylketen to carboxylic acids in ether solution, the
low-temperature fractionation. It was stored in vacuo a t
reaction being essentially first-order in keten and acid. -183 "C and the amount required for each run redistilled
The kinetics of the gas-phase reaction between keten and a t -78 "C before use.
acetic acid are reported on here,
EXPERIMENTAL RESULTS AND DISCUSSION
Apparatus and Procedure.-A conventional static appar- The gas-phase reaction between keten and acetic acid
atus was used. Dead spaces were heated to prevent has been studied in the temperature range 106-215 "C
condensation and the mean dead space correction was over keten pressures from 8-140 mmHg and acid
0.9%. The reaction was studied in unpacked and packed pressures 6-190 mmHg (1 mmHg = 133.28 N m-2).
silica vessels, the surface-to-volume ratios of which were Stoicheiometry.-Acetic anhydride is the only product,
0-88 and 9-69 cm-l respectively. Temperatures were dimerisation of keten being negligible under the condi-
measured by Chromel-Alumel thermocouples which had tions used. The stoicheiometry of reaction is accurately
been calibrated with a standard platinum/platinum- represented by equation (2) and Table 1 shows that one
rhodium couple. An electronic platinum resistance thermo-
meter controller maintained temperatures constant to
within & 0.2". Rates were measured in carbon-coated
CH&O + CH,*CO,H __c (CH,*CO),O
vessels, prepared by pyrolysing acetic acid in them a t molecule of keten reacts to form one molecule of
600 OC, and in uncoated vessels. anhydride with a pressure drop corresponding to two
Runs were carried out by adding required pressures of gaseous molecules forming one.
keten and acetic acid separately to the vessel. Premixing
was not feasible, since rates of reaction were appreciable a t TABLE1
temperatures needed to prevent acid condensation. Runs Stoicheiometry of the reaction of keten with acetic acid
were performed in which keten was added to acid and T/"C Ap/mmHg A(keten)/mmHg A(anhydride)/mmHg
vice uersa, and mean values taken. The difference between 170 -46.4 -47.2 +48.9
the two runs was typically 10%. Pressures were measured 181 -45.2 - 45.8 +44.8
with a glass spiral manometer. At the end of a run the 192 -45.3 -45.1 +46.9
contents of the vessel, which comprised keten, acetic acid, 192 -43.6 -43.7 +46.5
and acetic anhydride only, were passed through a packed
trap a t -95 "C. Keten was pumped over by diffusion and
Kinetics.-The reaction is homogeneous. Rates are
Toepler pumps, measured in a gas burette and by i.r. identical in carbon-coated and untreated vessels. Rates
absorption at 2150 cm-l. Acid and anhydride were then in the packed vessel are 20% lower than unpacked
diluted to 5 cm3 with acetic acid, and acetic anhydride vessels, but pressure-time curves are slightly sigmoid in
estimated by its U.V. absorption a t 256 nm. these vessels and the effect would seem to be due to the
Rate constants were calculated from initial rates and finite time taken for diffusional mixing of acid and keten
from second-order plots, the latter being more reliable when in the packed vessel, which is filled with glass tubes of
rates were fast. When keten and acid pressures were relatively small bore.
widely different normal second-order equations for a # b Reaction orders were measured at 120 and 157'
were employed. When nominal initial pressures were the
(Table 2). The order in keten is 1-00 &- 0.03 at both
same, the equation used depended on how similar were the
pressures experimentally obtained. If they were equal to temperatures. The order in acetic acid is more com-
within 0.4%, l / ( a - x ) was plotted against time, where plicated, since both monomer and dimer are present.
n = b. If pressures differed by from O . P - 5 % it was Order plots in total acid pressure are noticeably curved
necessary to use equation (1) where A = (a - b ) / ( a - x ) at 120 "C with a mean slope of 0.76. To obtain order
1 plots in acid monomer the equilibrium constant for
t ( a - b) dissociation of acetic acid dimer is needed. Published
values of the standard enthalpy and entropy of dissoci-
ation vary considerably; the mean of post-war values
* S. Glasstone, ' Textbook of Physical Chemistry,' 2nd edn.,
J . M. Rriody, P. J. Lillford, and D. P. N. Satchell, J . Chem. Macmillan, London, 1948, p. 1055.
SOC.( B ) ,1968, 885. 3 G . Allen and E. F. Caldin, Quad. Rev., 1953, 7, 261.
 View Article Online

1728 J. Chem. SOC. (B), 1971

TABLE2 dynamic data could alter the result by more than k0.1.
Reaction orders in keten and acid monomer * At 157 "C, where acid dimerisation is much less, the order
(a) Keten; p(tota1 acid) = 50 in total acid is 0.98 and in acid monomer 1-03.
103R0/p(keten) The reaction is thus second-order overall, first in
10*R0/mmHgs-l
p(keten) S-1 keten and first in acid monomer, and second-order plots
(i) T 120.5" 20.1 2.19 1.05 give good straight lines a t extents of reaction not near
49.1 4.99 1.02
92.0 10.51 1-14 equilibrium. The indication is either that keten does
103.3 11-65 1.13 not react with acid dimer or that the rate of any such
(ii) T 156.0 7.4 1.87 2.52
Published on 01 January 1971. Downloaded by University of Kentucky on 13/01/2016 22:18:30.

reaction is much slower than that with acid monomer.

25.9 6.17 2.38
50-2 12-2 2-43 Arrhenius plots yield the equation k = 105*81*0*05-
140.1 33.3 2.38 exp (- 46,260 &- 240/RT) 1 mol-l s-l, a t constant
(b) Acid monomer; p(keten) = 50 concentration. The kinetic results are consistent with a
1O3R0/p(acid
p(tota1 #(acid 1O2R0/ monomer)
six-centre cyclic transition-state formed by attachment
acid) monomer) mmHg s-l S-1 of the carbonyl oxygen of the acid to the carbonyl carbon
(i) T 120.5 8.1 7.6 1.67 2.19
14.8 13.4 2.53 1.88
19.0 16-7 2.93 1.76
20.5 17.9 3.04 1-70
117.5 73.5 12.1 1-65
(ii) T 157.5 6-8 6.7 1.85 2.76
9.5 9.4 2.48 2.64
21.5 20.9 5-84 2.79
50.2 46.8 12.8 2.74 of keten. This is of course the transition state postu-
101-5 89.3 28.3 3.16 lated for the reverse reaction, the decomposition of
* R, is initial rate, pressures in mmHg. acetic anhydride. The low A factor is consistent with a
bimolecular association of two non-linear molecules
is AH" = +604 kJ mol-l, AS" = +139 J K-l mol-l, through a cyclic transition state.5
giving K = 0.162 atm at 120 "C.
Using this value the order in acid monomer is 0.97 at We thank the S.R.C. for a maintenance grant (toH. H. D.).
120", but use of selected4 rather than mean thermo-
[l/669 Received, May 3rd, 19711
4 M. M. Davies, P. Jones, D. Patniak, and E. A. Moelwyn-
Hughes, J. Chem. SOC.,1961, 1249; M. D. Taylor, J. Amev. 6 S. W. Benson, ' T h e Foundations of Chemical Kinetics,'
Chem. SOC.,1951, 73, 315. McGraw-Hill, New York, 1960, p. 302.