PHYSICAL CHEMISTRY

A. M. BRODSKII, R. A. KALINENKO,
Corresponding Member of the Academy of Sciences of the USSR K. P. LAVROVSKII, L. V. SHEVELKOVA,
Yu. P. YAMPOLSKII

ON THE REGULARITIES OF THE TRANSFORMATIONS OF ETHYLENE AND ACETYLENE DURING HIGH-TEMPERATURE DECOMPOSITION OF HYDROCARBONS

Most published studies on the thermal decomposition of ethylene are devoted to the investigation of its transformations either at temperatures below 800°, or in shock waves at temperatures above 1000° (1–5). The temperature interval 800–1000° remains the least studied.

In connection with this, we carried out an investigation of the thermal decomposition of ethylene with additions of \(\mathrm{C_2H_2}\), labeled with radiocarbon \(\mathrm{C^{14}}\), in the temperature interval 800–1000° at a pressure of 100 mm Hg. Study of the distribution of radioactivity among the pyrolysis products of \(\mathrm{C_2H_4}\) made it possible to elucidate the pathways and regularities of the transformations of \(\mathrm{C_2H_2}\) under the conditions of thermal decomposition of \(\mathrm{C_2H_4}\). The experiments were conducted in a flow vacuum apparatus with a metallic reactor according to the procedure described in (6). The thermal decomposition reaction of ethylene was carried out in a boiling bed of heat-transfer medium, for which finely ground quartz was used. The experiments were performed under conditions of complete mixing of the reactants. The reaction products were analyzed chromatographically.

Two series of experiments were carried out on mixtures of the following compositions: 99.44% \(\mathrm{C_2H_4}\) + 0.56% \(\mathrm{C_2H_2}\) and 97.3% \(\mathrm{C_2H_4}\) + 2.7% \(\mathrm{C_2H_2}\). The specific radioactivities of \(\mathrm{C_2H_2}\) in the initial mixtures were: \(5.1 \cdot 10^6\) and \(1.1 \cdot 10^6\) imp/\(\mathrm{cm^3 \cdot min}\); the specific radioactivity of the entire mixture in both cases was \(285 \cdot 10^4\) imp/\(\mathrm{cm^3 \cdot min}\).

The reaction mixture was separated chromatographically into individual components, and then their specific radioactivity was determined with the aid of internal-fill counters. The specific radioactivity of the coke deposited during the reaction on the surface of the heat-transfer medium was determined by burning it in an oxygen atmosphere at 500–600°.

Thermal decomposition of ethylene under the experimental conditions proceeds (6) by a first-order law with rate constant \(k_1\), whose Arrhenius dependence in the temperature interval considered is as follows*:

\[ \lg k_1 = (11.35 \pm 0.2) - (60\,000 \pm 1500) / 4.58\,T\ \mathrm{sec}^{-1}. \tag{1} \]

The principal reaction products under the conditions studied are \(\mathrm{H_2}\), \(\mathrm{CH_4}\), \(\mathrm{C_2H_2}\), \(\mathrm{C_4H_6}\), \(\mathrm{C_6H_6}\), and coke. Ethane, propylene, allene, methylacetylene, vinylacetylene, toluene, and cyclopentadiene are obtained in considerably smaller amounts (6). Since in the present work the investigations were carried out in a reactor with complete mixing of the products, from the ratio of the specific radioactivities of the \(i\)-th products and \(\mathrm{C_2H_2}\) \((a_i/a_{\mathrm{C_2H_2}})\) it is possible directly to estimate the degree of participation of \(\mathrm{C_2H_2}\) in the formation of these products. Thus, if \(a_i = a_{\mathrm{C_2H_2}}\), then one molecule of \(\mathrm{C_2H_2}\) participates in the formation of the \(i\)-th product,

* It should be noted that the numerical values of the rate constants for the decomposition of pure ethylene and of ethylene in the system of cracked ethane (7) coincide within the limits of experimental error.

if \(a_i < a_{\mathrm{C_2H_2}}\), there are pathways for formation of the \(i\)-th product without participation of \(\mathrm{C_2H_2}\), etc. Figure 1 shows the ratios \(a_i/a_{\mathrm{C_2H_2}}\) at various temperatures and degrees of conversion. Since at low temperatures all \(a_i < a_{\mathrm{C_2H_2}}\) (except \(2a_{\text{coke}} \approx a_{\mathrm{C_2H_2}}\)), there are pathways for the formation of all products without participation of \(\mathrm{C_2H_2}\). With increasing temperature, degree of conversion,

Fig. 1. Dependence of \(a_i/a_{\mathrm{C_2H_2}}\) on temperature (in the case of methane and coke, the ordinate shows \(2a_{\mathrm{CH_4}}/a_{\mathrm{C_2H_2}}\) and \(2a_{\text{coke}}/a_{\mathrm{C_2H_2}}\)).
1 — methane (a); 2 — ethane (b) and ethylene (c); 3 — propylene (d); 4 — divinyl (e); 5 — vinylacetylene (f); 6 — cyclopentadiene (g); 7 — allene (h) and methylacetylene (i); 8 — benzene (k) and toluene (l)

and the content of \(\mathrm{C_2H_2}\) in the reaction mixture, all ratios \(a_i/a_{\mathrm{C_2H_2}}\), except \(2a_{\text{coke}}/a_{\mathrm{C_2H_2}}\) and \(2a_{\mathrm{CH_4}}/a_{\mathrm{C_2H_2}}\), increase monotonically. Throughout the temperature interval investigated, \(2a_{\mathrm{CH_4}}\), \(a_{\mathrm{C_2H_4}}\), \(a_{\mathrm{C_2H_6}}\), \(a_{\mathrm{C_3H}}\), and \(a_{\mathrm{C_4H_6}}\) remain substantially smaller than \(a_{\mathrm{C_2H_2}}\); \(a_{\mathrm{C_3H_4}}\) approach \(a_{\mathrm{C_2H_2}}\), the specific activity of cyclopentadiene \(a_{\mathrm{C_5H}}\) and vinylacetylene \(a_{\mathrm{C_4H_4}}\) approaches \(2a_{\mathrm{C_2H_2}}\), and that of benzene and toluene approaches \(3a_{\mathrm{C_2H_2}}\). The curves for the temperature dependence of \(2a_{\text{coke}}/a_{\mathrm{C_2H_2}}\) and \(2a_{\mathrm{CH_4}}/a_{\mathrm{C_2H_2}}\) have a minimum at \(900^\circ\). The fact that \(a_{\mathrm{C_6H_6}}/a_{\mathrm{C_2H_2}}\) varies within the range from 0 to \(\sim 3\) indicates that in the system under consideration at least two pathways for benzene formation occur at comparable rates: reactions involving only acetylene molecules and the products of its transformations (for example, \(\mathrm{C_2H_2 + C_4H_4 \to C_6H_6}\)), and reactions without participation of acetylene. Since traces of cyclohexene, which disappeared as the temperature was raised, were detected in the reaction products at temperatures below \(900^\circ\), this compound was apparently an intermediate in the formation of labeled benzene. In addition, the occurrence of reactions forming \(\mathrm{C_6H_6}\) with participation of both \(\mathrm{C_2H_2}\) and \(\mathrm{C_2H_4}\) molecules is not excluded.

Since over the entire temperature interval investigated \(\mathrm{C_2H_2}\) did not take a substantial part in the formation of \(\mathrm{CH_4}\) and \(\mathrm{C_4H_6}\), the corresponding rate constants for formation of these products in the thermal decomposition of \(\mathrm{C_2H_4}\) may be calculated from the formulas*

\[ [\mathrm{CH_4}] = k_2[\mathrm{C_2H_4}]\tau; \tag{2} \]

\[ [\mathrm{C_4H_6}] = k_3[\mathrm{C_2H_4}]^2 \tau/(1+k_4\tau). \tag{3} \]

* Earlier, in the cracking of ethane with additions of labeled ethylene, it was shown \((^7)\) that \(a_{\mathrm{C_4H_6}}/a_{\mathrm{C_2H_4}} = 2\), i.e., that divinyl is formed by the overall reaction of two ethylene molecules. Since the kinetic curve for accumulation of divinyl has a maximum and the quantity \([\mathrm{C_4H_6}]/[\mathrm{C_2H_4}]^2\tau\) decreases with increasing \(\tau\), under the experimental conditions relatively fast reactions of thermal decomposition of divinyl occur, proceeding by first order with the rate constant \(k_4\) \((^6)\).

The Arrhenius dependences of the constants \(k_2\) and \(k_3\) (Fig. 2) can be represented as follows:

\[ \lg k_2=(11.75\pm 2)-(67000\pm 10000)/4.58\,T\quad (\mathrm{sec}^{-1}), \tag{4} \]

\[ \lg k_3=(-6.7\pm 2)-(63000\pm 10000)/4.58\,T\quad (\mathrm{molecule}^{-1}\cdot \mathrm{cm}^3\cdot \mathrm{sec}^{-1}). \tag{5} \]

Acetylene is unstable under the conditions studied and reacts with the formation of various products. In this process the main part of the radioactivity of \(C_2H_2\) (from 0.5 to 0.75 of the initial radioactivity) passes into coke; benzene, vinylacetylene, ethylene, and methane (Table 1). The rate constant of the overall first-order transformation of \(C_2H_2\), \(k_d\), can be calculated from data on the consumption of labeled \(C_2H_2\):

\[ A^0_{C_2H_2}/aA_{C_2H_2}=1+k_d\tau, \tag{6} \]

where \(A^0_{C_2H_2}\) and \(A_{C_2H_2}\) are the concentrations of labeled \(C_2H_2\) in the initial mixture and in the reaction products; the quantity \(A_{C_2H_2}\) is equal to \(a_{C_2H_2}\) multiplied by the fraction of \(C_2H_2\) in the reaction products; \(k_d=\sum k_i C_i\). It is very interesting that the curve of the dependence of \(\lg k_d\) on \(1/T\) has a characteristic minimum at 900° and cannot be represented by the Arrhenius formula (Fig. 3). The rate constant of the overall decomposition of acetylene, \(k_d\), can be expressed as the sum \(k_d=k_c^*+k_g\), where \(k_c^*\) is the rate constant for the formation of coke from acetylene, determined by the formula

\[ k_c^*=A_{\mathrm{coke}}/A_{C_2H_2}\tau, \tag{7} \]

and \(k_g\) is the rate constant for the formation from acetylene of the remaining gaseous products. The course of the curve of the dependence of \(\lg k_c^*\) on \(1/T\) also does not correspond to the Arrhenius formula and has a minimum at 900° (Fig. 3). The value \(k_g\), on the contrary, changes monotonically with temperature, with an apparent activation energy of \(\sim 10\) kcal/mole (Fig. 3).

Fig. 2. Dependence of \(\lg k_3\) and \(\lg k_2\) on \(1/T\).

Table 1

Conditions of the experiments and concentration of labeled reaction products \(A_i\)

* Values of \(A_i\) are found from the formula \(A_i=\dfrac{a_i}{7.37}C_i\), where \(C_i\) is the percentage concentration of the \(i\)-th component of the mixture.
* As a blank experiment showed, during chromatographic separation of the reaction products, 0.03% of radioactive acetylene is entrained by ethylene. In this connection the concentrations of labeled ethylene are somewhat overestimated.
** Experiment with addition of 2.7% \(C_2H_2\). In the remaining experiments, the addition was 0.56% \(C_2H_2\).

The results of this work show that acetylene plays a very important role in coke-formation reactions during high-temperature transformations of hydrocarbons. At the highest and lowest of the investigated temperatures, the ratio \(2a_{\text{coke}}/a_{\mathrm{C_2H_2}}\) is close to unity, i.e., the greater part of the carbonaceous deposits is formed as a result of acetylene transformations. At the same time, since at low temperatures all \(a_i\) are small, it should be assumed that under these conditions coke is formed during decomposition, polymerization, or condensation directly from \(\mathrm{C_2H_2}\). At temperatures of 950–1000°, the ratio \(2a_{\text{coke}}/a_{\mathrm{C_2H_2}} = 0.5\). Thus, under these conditions a substantial part of the carbonaceous deposits is formed not through acetylene, but as a result of transformations of ethylene and other hydrocarbons having low specific radioactivities. The apparent rate constant for the formation of nonradioactive coke, \(k_c^0\), calculated from the formula

Fig. 3. Dependence of \(\lg k_d\), \(\lg k_c^*\), and \(\lg k_g\) on \(1/T\)

\[ k_c^0 = (1 - 2a_{\text{coke}}/a_{\mathrm{C_2H_2}})[\text{coke}]/[\mathrm{C_2H_4}]\tau, \tag{8} \]

has an activation energy of about 80 kcal/mole. Such a high value of the activation energy indicates that coke is formed as a result of decomposition reactions. The numerical value of \(k_c^0\) at temperatures of 950–1000° is 2–3 times smaller than \(k_c^*\).

Since in the present work and in (6) complete data have been obtained on the formation and decomposition of acetylene under the conditions studied, one may attempt to determine the rate constant for the formation of \(\mathrm{C_2H_2}\) from ethylene, \(k_5'\), taking its decomposition into account:

\[ [\mathrm{C_2H_2}] = k_5'[\mathrm{C_2H_4}]\tau/(1+k_d\tau). \tag{9} \]

However, the value \(k_5'\), calculated by formula (9), does not remain constant with variation in the extent of conversion. At the same time, the value \(k_5\), calculated by the formula

\[ [\mathrm{C_2H_2}] = k_5[\mathrm{C_2H_4}]\tau \tag{10} \]

without taking into account the subsequent decomposition of acetylene, is constant within the experimental error, and its dependence on temperature is expressed by the Arrhenius formula

\[ \lg k_5 = (15.34 \pm 1.0) - (86\,000 \pm 10\,000)/4.58\,T\;(\text{sec}^{-1}). \tag{11} \]

Thus, it turns out that \(\mathrm{C_2H_2}\) added to the initial \(\mathrm{C_2H_4}\) decomposes at a considerably higher rate than the \(\mathrm{C_2H_2}\) formed in the course of the reaction. This very interesting circumstance, the full explanation of which requires further investigation, may be connected with the formation in the reaction process of an excited triplet state of \(\mathrm{C_2H_2}\) (4).

Institute of Petrochemical Synthesis
named after A. V. Topchiev
Academy of Sciences of the USSR

Received
16 X 1964

CITED LITERATURE

1. V. Kevorkian, High Temperature Chemistry of the Light Hydrocarbons, Advances in Petroleum Chemistry, 5, 1962.
2. G. B. Skinner, E. Sokolovsky, J. Phys. Chem., 64, 1028 (1960).
3. G. I. Kozlov, V. G. Knorre, Kinetika i kataliz, 4, 189 (1963).
4. H. P. Palmer, F. L. Dormish, J. Phys. Chem., 68, 1553 (1964).
5. G. Dahlgren, J. E. Douglas, J. Am. Chem. Soc., 80, 5108 (1958).
6. Yu. P. Yampolsky, A. M. Brodsky et al., Neftekhimiya, 4, No. 5, 691 (1964).
7. A. M. Brodsky, R. A. Kalinenko et al., Kinetika i kataliz, 5, 49 (1964).
8. L. V. Shevelkova, A. M. Brodsky et al., DAN, 160, No. 2 (1964).