J. Phys. Chem. 1995,99, 13168-13172
13168
Thermal Unimolecular Reaction of Pyruvonitrile: Experimental and Computational Study on the Occurrence of Isomerization Kazumasa Okada and KO Saito* Department of Chemistry, Faculty of Science, Hiroshima University, Higashi-Hiroshima 724, Japan Received: April 4, 1995; In Final Form: June 27, 1995@
The thermal unimolecular reaction of pyruvonitrile diluted in Ar has been studied behind reflected shock waves over a temperature range 1014-1300 K, with a total density of ca. 1.0 x mol/cm3. The reaction was monitored by means of time-resolved vacuum-UV absorption and IR emission. Also, ab initio molecular orbital (MO) calculations were carried out in order to formulate a reaction mechanism. From the gas chromatographic analysis of the shock-heated gas and from ab initio MO calculations, it is suggested that the isomerization to acetyl isocyanide occurs under the present experimental conditions. It is estimated that the difference of enthaipies between these species is 5 kcal/moi. The first-order rate constants are expressed as kf = exp(-49.3 kcal mol-'/RT) s-! for the isomerization of pyruvonitrile to acetyl isocyanide and k, exp(-44.3 kcal mol-'/RT) s - I for the reverse reaction.
1. Introduction Previously we have investigated gas-phase thermal unimolecular reactions of relatively large molecules such as esters, ethers, and carboxylic acids.'-3 From these studies we have found that in the consecutive unimolecular process, there occurs thermally unfavorable reaction channel(s). We proposed that this unusual phenomenon can be explained on the basis of the dynamic behavior of the reacting molecule. Recently, we have studied the thermal unimolecular decomposition of acetaldoxime where methyl isocyanide is produced as well as acetonitrile. A classical trajectory calculation revealed the possibility of the production of a thermally unstable p r o d ~ c t .In ~ the present study we consider pyruvonitrile as a candidate having the possibility of producing acetonitrile or methyl isocyanide during its thermal decomposition. Three reaction channels are considered for the thermal reaction of pyruvonitrile: CH,COCN
-
CH,COCN
CO 4-CH,CN(CH,NC)
-
CH,COCN
CH,=CO
-
+ HCN
CH,CONC
(1)
(2) (3)
In the past, few studies reported this gas-phase reaction. Bennett et aL5 investigated the pyrolysis at 470 "C in a flow system. They reported that the reaction proceeds via channels 1 and 2 competitively. At the present stage, however, there are no kinetic data for this reaction. Therefore, we first need to investigate the kinetics of this thermal reaction both experimentally and theoretically. 2. Experimental Section
The experiments were performed in a pressure-driven shock tube made of stainless steel. A schematic diagram of our experimental apparatus is shown in Figure 1. The apparatus and procedures used in the present study are the same as that have been described previously,6 hence only a brief description of the system is given here. The driven section is 3.67 m long @
Abstract published in Advance ACS Absrrucfs, August 1, 1995.
0022-365419512099-13168$09.00/0
TRIG
1 oscilloscope
Ampinier
D C +12V H V ;wOV
Figure 1. Schematic diagram of the apparatus.
with a 9.4-cm i.d. and is evacuated by a 6-in. oil-diffusion pump to less than 2 x Torr before each run. It is separated from the driver by a polyester diaphragm. Shock waves are generated by bursting the diaphragm with a needle. Three pressure transducers are mounted flush with the inside wall 160 mm apart near the end of the driven section. Two of them are used for the measurement of the incident shock speed which is determined by counting the time intervals of shock-arrival signals with a universal counter (Takeda Riken, TR-5104G) with an accuracy of 0.1 ps. All experiments were conducted behind reflected shock waves. Reflected shock conditions for each run were calculated from the measured incident shock speed using the shock relations for the ideal gas. A pair of MgF2 windows is mounted on the tube walls 2 cm upstream from the end plate. The reaction is monitored by observing its vacuum-UV absorption through these windows. A microwave discharge tube containing flowing He with a few percent of additional gas is used to generate the atomic resonance light. The wavelength is selected by a monochromator (Minuteman Laboratories Inc., 302VM) and the vacuumUV light intensity is detected by a photomultiplier (Hamamatsu, R431s). The output signals are fed into a digital storage oscilloscope and subsequently analyzed to obtain kinetic parameters. Preliminary experiments were carried out in order 0 1995 American Chemical Society
J. Phys. Chem., Vol. 99, No. 35, 1995 13169
Thermal Unimolecular Reaction of Pyruvonitrile
-100 r
I
tu
'
1
I
I
RTS.
0 I mol% pyruvcmtrile
T=1236K
0
0
200 400 600 time / ps
Figure 2. A typical absorption profile at 174.4 nm. Conditions: T5 = 1236 K, e5= 1.090 x moYcm3, 0.10 mol 8 reactant in Ar. IS and RS indicate the incident and the reflected shock fronts, respectively.
to find a suitable wavelength for rate analysis, and the observations at 174.4 nm were found to be appropriate for the present reaction system. It was shown that the absorbance of the reactant is half that of the products at this wavelength. Timeresolved measurements of the reaction were also performed by using the IR emission from the product. The IR radiation is passed through a band-pass interference filter (4.90 pm) and is detected by a photoconductive HgCdTe element cooled at 77 K. The recording system is the same as in the absorption experiment. The pyruvonitrile for the experiments was obtained commercially. After the first fraction was pumped off in a vacuum line, the sample was expanded into an evacuated glass flask, diluted with Ar (99.9995% purity) to 0.10-0.20 mol % and stored. For the purpose of qualitative analysis of the shock-heated gas component, the gas was taken into a glass cylinder and analyzed by gas chromatography on a 2 m Porapak-R column.
TABLE 1: Rate Data for the Pyruvonitrile Reaction PJTorr MS 1 O S g ~ ~ / ( m o ~ c m 3 Tp/K ) ktst/s-i 40.1 40.6 40.0 39.0 40.0 40.5 40.0 40.0 40.0 39.2 40.0 39.8 40.0 40.0 40.0 40.0 40.0 40.0 40.0 41.6 40.0 40.0 40.0
0.10 mol % Pyruvonitrile in Ar 2.079 0.9925 2.254 1.113 2.237 1.090 2.208 1.044 2.274 1.100 2.179 1.070 0.9850 2.054 1.017 2.117 1.009 2.110 0.9851 2.126 1.005 2.109 0.9492 2.034 1.074 2.221 1.038 2.162 0.9434 1.999 1.022 2.125 1.026 2.121 1.079 2.239 1.062 2.213 1.083 2.162 1.011 2.129 2.076 0.9752 0.9705 2.064
(174 nm) 1091 1258 1236 1212 1286 1179 1057 1123 1121 1151 1123 1058 1229 1172 1014 1131 1120 1254 1229 1169 1149 1103 1089
1.88 x 8.17 x 7.43 x 4.91 x 6.03 x 1.79 x 2.81 x 4.58 x 4.37 x 1.17 x 8.90 x 2.32 x 3.27 1.20 x 5.73 x 7.06 x 6.77 x 3.04 x 3.34 x 1.23 x 8.10 x 5.19 x 2.55 x
lo2 10) 103 103 103 103 lo2 lo2 102 103 lo2 lo2 103 103 101 lo2 lo2 103 103 10; lo2 lo2 lo2
40.0 40.0 40.0 40.0 40.0 40.0
0.20 mol Ti Pyruvonitrile in Ar 2.135 1.01 1 2.236 1.071 2.287 1.107 2.218 1.069 2.223 1.083 2.288 1.120
(4.90 pm) 1159 1256 1300 1230 1223 1287
1.90 x 4.25 x 6.60 x 2.77 x 3.65 x 1.13 x
lo3 lo3 lo3 lo3 lo3 lo4
Quantities with the subscript 5 refer to the thermodynamic state of the gas in the reflected shock region.
1
0.20mol% pyruvonitriie
T-1300 K
l-TrT-TT----r
3. Computational Methods Ab initio MO calculations were carried out by using the GAUSSIAN 88' and GAUSSIAN 908 program packages. The structures of the stationary points including transition states were fully optimized at the Hartree-Fock level by using the energy gradient technique. Vibrational frequencies were calculated at the HF/3-21G level using analytical second derivatives to correct for the zero-point energies. Finally, conventional transitionstate theory9 was used to estimate the rate constants.
4. Results and Discussion A. Experimental Results. The experiments were performed behind reflected shocks over the temperature range 1014- 1300 K and the total density range of (0.94-1.1) x mol/cm3. Time-dependent behavior was observed via both vacuum-UV absorption and IR emission during the shock heating time. Gas chromatographic analysis of the shock-heated gas revealed an absence of reaction products at temperatures
