The kinetics of gaseous oxidation reactions
<p>There are many interesting problems associated with the thermal oxidation of hydrocarbons. The relationship between rate and structure is remarkable, the temperature dependence is anomalous and the kinetics are both complex and not fully understood. There is often a long induction period as...
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
1959
|
| _version_ | 1817931657035907072 |
|---|---|
| author | Horscroft, R |
| author_facet | Horscroft, R |
| author_sort | Horscroft, R |
| collection | OXFORD |
| description | <p>There are many interesting problems associated with the thermal oxidation of hydrocarbons. The relationship between rate and structure is remarkable, the temperature dependence is anomalous and the kinetics are both complex and not fully understood. There is often a long induction period associated with the development of branching chains, during which time intermediates, such as peroxides and aldehydes, are built up in small quantity. The reaction is also very much subject to the influence of small homogeneous and heterogeneous additions. This means that not only can ignition be promoted, but by suitable means delayed or prevented, a subject of very great practical importance.</p> <p>The objective of the work here described has been to investigate the gas phase oxidation of a series of cycloparaffins and to compare the kinetics with similar open chain systems. Particular reference has been made to the problems of the rate-structure relationship and the effect of surfaces, both on the induction period and on the rate. The systems studied for this purpose have been a aeries of cycloparaffins from cyclopropane to cyolooctane, excluding cyclobutane but including various methyl substituted cyclopentanes and hexanes. The more detailed study of the kinetics and of the relevant inhibition phenomena been restricted to cyclopentane, n-pentane being used to provide comparison with the open chain paraffins.</p> <p>The work was carried out to a conventional static apparatus at pressures from 100 to 700 mm Hg, reaction vessels being of pyrox or silica. The course of the reaction was followed by pressure increase, measured on a mercury manometer. Product analysis by mass spectrometer, gas chromatography and chemical methods showed that pressure changes were a good measure of the extent of the reaction. Oxidation reactions in the gas phase have an anomalous temperature dependence resulting in "high" and "low" temperature regions of of reaction. Most of the present investigation has been carried out in the low temperature range, below 500°C, with a little comparative work done at higher temperatures.</p> <p>It was first considered desirable to make some comparison of n-pentane and cyclopentane, since the oxidation of the normal paraffins has already been investigated, and the comparison would show how far results could be regarded as general. These experiments were carried out at 230°C, with variation of both the hydrocarbon and oxygen pressures. In general the kinetics of the two paraffins are very similar. The reciprocal induction period was in all cases linearly proportional to the pressure of hydrocarbon or oxygen, the dependence being a little steeper with increasing hydrocarbon pressure than oxygen pressure. The dependence upon reactant pressure of the rate was found to be more complex. There is a limiting pressure of either hydrocarbon or oxygen above which increase in the pressure of the other reactant causes a rapid rise in rate to explosion. Below this critical pressure the rate attains a limiting value with increase either in oxygen or in hydrocarbon. Thus under certain conditions the rate becomes independant of the concentration of one reactant. The change from one type of behaviour to the other is gradual, and there is a region where the rate is linearly proportional to the concentration of one reactant over a wide pressure range. This type of behaviour was observed with several hydrocarbons. It had been previously supposed that with n-pentane the rate attained a limiting value with increase of oxygen, but it now appears that this is not always so.</p> <p>The comparison of a series of cycloparaffins was made at 250°C (except cyclopropane which was investigated at a higher temperature). The rate increases with the size of the ring, in the same way as it does in the open chain paraffins with length of the chain. The only anomaly observed was with cyclooctane, which oxidised at a rate between cyclopentane and cyclohexane. Methyl groups on the ring, as in branched open chain paraffins, progressively lower the rate.</p> <p>The effect of increasing the surface to volume ratio of the reaction vessel, was investigated. Fourfold and sixfold increases lengthened the induction period by a small amount and decreased the rate, besides raising the limiting pressures mentioned above. Much higher pressures ware attained before the rate accelerated to explosion, and in some cases a limiting rate replaced an explosion.</p> <p>The result of various small homogeneous additions was next investigated. Peroxides, aldehydes and some halogens all shorten the induction period and in most cases increase the rate. This is thought to be due to the rapid production of radicals which initiate the main reaction. Isopropyl halides had little or no affect on the induction period or rate, aniline, carbon tetrachloride and carbon tetrafluoride had none. Inert gases in the form of nitrogen and carbon dioxide had no effect, even in large amounts.</p> <p>Experimental evidence suggests that the initiation reaction may take place largely on the surface of the containing vessel. In order to try and utilise this fact for the suppression of the reaction, various surface treatments were investigated. These consisted in coating the inside of the reaction vessel with a thick layer of various salts, deposited from an aqueous solution. Those used were a series of alkali halides and some salts of di- and tri- valent metals. In the first instance experiments on the length of the induction period were carried out between 250°C and 290°C. All the salts caused a increase in the induction period (θ) and Arrhenius plots of log 1/θ against 1/T<sub>abs.</sub> gave straight lines. The magnitude of this effect was large, induction periods being increased by up to a thousandfold. Some investigations were made on the rate of reaction at 250°C and on account of the length of the induction period at this temperature, a trace of ditertiary butylperoxide was used as an initiator. Only a few of the less efficient salts gave a measurable rate, which was from 10% to 50% of the normal value. The others gave no measurable pressure increase, even with large amounts of initiator. It was further discovered that in the case of at least two salts, a few grams of the loose salt in a clean pyrex vessel had the same inhibiting effect as coating the whole interior of the vessel. This type of inhibition is not confined to salts, since silver wire, platinum wire and gold foil also inhibit the reaction.</p> <p>The temperature coefficient of the reaction was next investigated with cyclopentane in both salt-coated and uncoated vessels, and is anomalous. Two sections of smooth reaction are separated by a temperature region of cool flames and explosion. On raising the temperature the rate first increases and then gives way to explosion: this is succeeded by cool flames which as the high temperature region is approached give way to a smooth reaction with a negative temperature dependence. Eventually this becomes positive again and the rate finally rises to explosion. The whole process occurs over a temperature rang* of about 250°C. The induction periods show no anomalous behaviour, and give one Arrhenius plot through both high and low temperature with no break in the cool flame portion. Similar experiments were carried out with salt-coated vessels. With lithium bromide the rate in the high temperature region was little affected, but at lower temperatures where cool flames might be expected, a very marked reduction in rate was noticed. Magnesium sulphate caused a large reduction in rate in the high temperature region.</p> <p>The principal reaction products were carbon monoxide and water with some carbon dioxide, others being in very low concentrations, Other products and intermediates detected were formaldehyde, acetaldehyde, n-butyric aldehyde, butene and an unidentified peroxide or peroxides. The concentration of the principal products followed the rate of pressure rise closely. The peroxide concentration (determined by an iodometric method) passed through a well defined maximum at the time of maximum rate. In the cool flame and explosion region the peroxide concentration rose with the pressure increase and immediately after the explosion fell back to a lower value. In salt-coated vessels the peroxide concentration was most markedly reduced, both at low and high temperatures.</p> <p>The interpretation of these results will now be considered. The kinetics of the cycloparaffins seem to be analogous to those of the n-paraffin series. The reactants can apparently both start and end chains, since under certain conditions the rate can become independent of the concentration of one reactant, and even be depressed with increasing oxygen pressure. The general shape of the rate/pressure curves can be given by a simple empirical expression of the form:-</p> <p><em>[For the equation omitted here, please consult the PDF.]</em></p> <p>It is suggested that the progressive change in mode of reaction from low to high temperature occurs thorough the increased rate of of RO<sub>2</sub> radicals thus :-</p> <table> <tr><th>Low</th><th>High</th></tr> <tr><td>RH + O<sub>2</sub> → R + HO<sub>2</sub></td><td>RH + O<sub>2</sub> → R + HO<sub>2</sub></td></tr> <tr><td>R + O<sub>2</sub> → RO<sub>2</sub></td><td>R + O<sub>2</sub> → RO<sub>2</sub></td></tr> <tr><td>RO<sub>2</sub> + RH → ROOH + R</td><td>RO<sub>2</sub> → aldehydes → products</td></tr> <tr><td>ROOH → branching → products</td></tr> </table> <p>The cool flames are probably formed by the decomposition of the peroxide. The peroxide found experimentally at high temperatures is possibly hydrogen peroxide and not an organic peroxide. The peroxide postulated at low temperature may well be a cyclic one and the cyclic peroxy radicals of the high temperature mechanism would split to give one free radical and probably an aldehyde. The fact that the ring has to be broken means that there is a fundamental difference from the n-paraffins at some stage, although the resulting kinetics seem to be the same. The anomalous form of the temperature coefficient is typical of the oxidation of higher hydrocarbons and shows that two separate mechanisms oust exist, one showing a <em>decrease</em> in rate as the temperature is increased over a certain range.</p> <p>The structural effects in the cycloparaffins are again very similar to those in the n-paraffin series, methyl groups strongly decreasing the rate. This is probably due, either to the stabilising influence they have on a peroxide intermediate, or to the production of some more stable intermediate from the easily attacked tertiary carbon atoms.</p> <p>Increase in the surface/volume ratio of the reaction vessel indicates that chain ending on the walls can control the reaction to a large extent. The addition of almost any material which can produce free radicals under the conditions of the experiment, (e.g. aldehyde or peroxide) reduces the length of the induction period and in large amounts raises the rate. Evidently there are many free radicals capable of initiating the chain and if these are produced over a period of time, as in the oxidation of an added aldehyde, their concentration can become large enough to upset the kinetic balance of the reaction and so raise the total rate.</p> <p>The effect of a salt coating on the reaction vessel walls is of great interest. It seems very unlikely that the initiation reaction is interfered with in any my, since both metal wires and loose salt are as efficient as a complete coating. The inhibition must occur through the destruction of an intermediate, possibly a peroxide, since this has a half life time longer than that of the transient free radicals. Salt and metal surfaces might, however, destroy assail radicals of the type HO<sub>2</sub>. Peroxide removal would explain the fact that the efficiency falls off as the temperature is raised towards the high temperature region.</p> <p>The analytical results having shorn that the rate of pressure change is a good measure of the extent of reaction and that the concentration of intermediates is always very small, the cycloparaffin ring must break to give radicals or fragments that are very rapidly converted to final products.</p> <p>From the point of view of the control or prevention of ignition the most significant results may be summarised as follows. Rapid reaction depends upon the slow build up of active intermediates. These are susceptible, in principle, to destruction both by homogeneous additions and by surface action. Homogeneous additions of an effective kind have proved difficult to find. On the other hand a considerable variety of surfaces exert a most marked effect in prolonging the induction period, in some cases almost indefinitely. The action is not improbably connected with the destruction of the intermediate peroxides. The chains by which reaction is propagated almost certainly are initiated on the vessel walls but prevention of this initiation does not seem to be the method of action of the inhibitors. Therefore the control of ignition by salts would be expected to be most effective when the salt is widely distributed thorough the gas. These considerations perhaps indicate the best lines of future work.</p> |
| first_indexed | 2024-03-06T18:07:48Z |
| format | Thesis |
| id | oxford-uuid:020373ce-eddb-4841-827a-b6bba10e1915 |
| institution | University of Oxford |
| last_indexed | 2024-12-09T03:25:30Z |
| publishDate | 1959 |
| record_format | dspace |
| spelling | oxford-uuid:020373ce-eddb-4841-827a-b6bba10e19152024-12-01T09:06:11ZThe kinetics of gaseous oxidation reactionsThesishttp://purl.org/coar/resource_type/c_db06uuid:020373ce-eddb-4841-827a-b6bba10e1915Polonsky Theses Digitisation Project1959Horscroft, R<p>There are many interesting problems associated with the thermal oxidation of hydrocarbons. The relationship between rate and structure is remarkable, the temperature dependence is anomalous and the kinetics are both complex and not fully understood. There is often a long induction period associated with the development of branching chains, during which time intermediates, such as peroxides and aldehydes, are built up in small quantity. The reaction is also very much subject to the influence of small homogeneous and heterogeneous additions. This means that not only can ignition be promoted, but by suitable means delayed or prevented, a subject of very great practical importance.</p> <p>The objective of the work here described has been to investigate the gas phase oxidation of a series of cycloparaffins and to compare the kinetics with similar open chain systems. Particular reference has been made to the problems of the rate-structure relationship and the effect of surfaces, both on the induction period and on the rate. The systems studied for this purpose have been a aeries of cycloparaffins from cyclopropane to cyolooctane, excluding cyclobutane but including various methyl substituted cyclopentanes and hexanes. The more detailed study of the kinetics and of the relevant inhibition phenomena been restricted to cyclopentane, n-pentane being used to provide comparison with the open chain paraffins.</p> <p>The work was carried out to a conventional static apparatus at pressures from 100 to 700 mm Hg, reaction vessels being of pyrox or silica. The course of the reaction was followed by pressure increase, measured on a mercury manometer. Product analysis by mass spectrometer, gas chromatography and chemical methods showed that pressure changes were a good measure of the extent of the reaction. Oxidation reactions in the gas phase have an anomalous temperature dependence resulting in "high" and "low" temperature regions of of reaction. Most of the present investigation has been carried out in the low temperature range, below 500°C, with a little comparative work done at higher temperatures.</p> <p>It was first considered desirable to make some comparison of n-pentane and cyclopentane, since the oxidation of the normal paraffins has already been investigated, and the comparison would show how far results could be regarded as general. These experiments were carried out at 230°C, with variation of both the hydrocarbon and oxygen pressures. In general the kinetics of the two paraffins are very similar. The reciprocal induction period was in all cases linearly proportional to the pressure of hydrocarbon or oxygen, the dependence being a little steeper with increasing hydrocarbon pressure than oxygen pressure. The dependence upon reactant pressure of the rate was found to be more complex. There is a limiting pressure of either hydrocarbon or oxygen above which increase in the pressure of the other reactant causes a rapid rise in rate to explosion. Below this critical pressure the rate attains a limiting value with increase either in oxygen or in hydrocarbon. Thus under certain conditions the rate becomes independant of the concentration of one reactant. The change from one type of behaviour to the other is gradual, and there is a region where the rate is linearly proportional to the concentration of one reactant over a wide pressure range. This type of behaviour was observed with several hydrocarbons. It had been previously supposed that with n-pentane the rate attained a limiting value with increase of oxygen, but it now appears that this is not always so.</p> <p>The comparison of a series of cycloparaffins was made at 250°C (except cyclopropane which was investigated at a higher temperature). The rate increases with the size of the ring, in the same way as it does in the open chain paraffins with length of the chain. The only anomaly observed was with cyclooctane, which oxidised at a rate between cyclopentane and cyclohexane. Methyl groups on the ring, as in branched open chain paraffins, progressively lower the rate.</p> <p>The effect of increasing the surface to volume ratio of the reaction vessel, was investigated. Fourfold and sixfold increases lengthened the induction period by a small amount and decreased the rate, besides raising the limiting pressures mentioned above. Much higher pressures ware attained before the rate accelerated to explosion, and in some cases a limiting rate replaced an explosion.</p> <p>The result of various small homogeneous additions was next investigated. Peroxides, aldehydes and some halogens all shorten the induction period and in most cases increase the rate. This is thought to be due to the rapid production of radicals which initiate the main reaction. Isopropyl halides had little or no affect on the induction period or rate, aniline, carbon tetrachloride and carbon tetrafluoride had none. Inert gases in the form of nitrogen and carbon dioxide had no effect, even in large amounts.</p> <p>Experimental evidence suggests that the initiation reaction may take place largely on the surface of the containing vessel. In order to try and utilise this fact for the suppression of the reaction, various surface treatments were investigated. These consisted in coating the inside of the reaction vessel with a thick layer of various salts, deposited from an aqueous solution. Those used were a series of alkali halides and some salts of di- and tri- valent metals. In the first instance experiments on the length of the induction period were carried out between 250°C and 290°C. All the salts caused a increase in the induction period (θ) and Arrhenius plots of log 1/θ against 1/T<sub>abs.</sub> gave straight lines. The magnitude of this effect was large, induction periods being increased by up to a thousandfold. Some investigations were made on the rate of reaction at 250°C and on account of the length of the induction period at this temperature, a trace of ditertiary butylperoxide was used as an initiator. Only a few of the less efficient salts gave a measurable rate, which was from 10% to 50% of the normal value. The others gave no measurable pressure increase, even with large amounts of initiator. It was further discovered that in the case of at least two salts, a few grams of the loose salt in a clean pyrex vessel had the same inhibiting effect as coating the whole interior of the vessel. This type of inhibition is not confined to salts, since silver wire, platinum wire and gold foil also inhibit the reaction.</p> <p>The temperature coefficient of the reaction was next investigated with cyclopentane in both salt-coated and uncoated vessels, and is anomalous. Two sections of smooth reaction are separated by a temperature region of cool flames and explosion. On raising the temperature the rate first increases and then gives way to explosion: this is succeeded by cool flames which as the high temperature region is approached give way to a smooth reaction with a negative temperature dependence. Eventually this becomes positive again and the rate finally rises to explosion. The whole process occurs over a temperature rang* of about 250°C. The induction periods show no anomalous behaviour, and give one Arrhenius plot through both high and low temperature with no break in the cool flame portion. Similar experiments were carried out with salt-coated vessels. With lithium bromide the rate in the high temperature region was little affected, but at lower temperatures where cool flames might be expected, a very marked reduction in rate was noticed. Magnesium sulphate caused a large reduction in rate in the high temperature region.</p> <p>The principal reaction products were carbon monoxide and water with some carbon dioxide, others being in very low concentrations, Other products and intermediates detected were formaldehyde, acetaldehyde, n-butyric aldehyde, butene and an unidentified peroxide or peroxides. The concentration of the principal products followed the rate of pressure rise closely. The peroxide concentration (determined by an iodometric method) passed through a well defined maximum at the time of maximum rate. In the cool flame and explosion region the peroxide concentration rose with the pressure increase and immediately after the explosion fell back to a lower value. In salt-coated vessels the peroxide concentration was most markedly reduced, both at low and high temperatures.</p> <p>The interpretation of these results will now be considered. The kinetics of the cycloparaffins seem to be analogous to those of the n-paraffin series. The reactants can apparently both start and end chains, since under certain conditions the rate can become independent of the concentration of one reactant, and even be depressed with increasing oxygen pressure. The general shape of the rate/pressure curves can be given by a simple empirical expression of the form:-</p> <p><em>[For the equation omitted here, please consult the PDF.]</em></p> <p>It is suggested that the progressive change in mode of reaction from low to high temperature occurs thorough the increased rate of of RO<sub>2</sub> radicals thus :-</p> <table> <tr><th>Low</th><th>High</th></tr> <tr><td>RH + O<sub>2</sub> → R + HO<sub>2</sub></td><td>RH + O<sub>2</sub> → R + HO<sub>2</sub></td></tr> <tr><td>R + O<sub>2</sub> → RO<sub>2</sub></td><td>R + O<sub>2</sub> → RO<sub>2</sub></td></tr> <tr><td>RO<sub>2</sub> + RH → ROOH + R</td><td>RO<sub>2</sub> → aldehydes → products</td></tr> <tr><td>ROOH → branching → products</td></tr> </table> <p>The cool flames are probably formed by the decomposition of the peroxide. The peroxide found experimentally at high temperatures is possibly hydrogen peroxide and not an organic peroxide. The peroxide postulated at low temperature may well be a cyclic one and the cyclic peroxy radicals of the high temperature mechanism would split to give one free radical and probably an aldehyde. The fact that the ring has to be broken means that there is a fundamental difference from the n-paraffins at some stage, although the resulting kinetics seem to be the same. The anomalous form of the temperature coefficient is typical of the oxidation of higher hydrocarbons and shows that two separate mechanisms oust exist, one showing a <em>decrease</em> in rate as the temperature is increased over a certain range.</p> <p>The structural effects in the cycloparaffins are again very similar to those in the n-paraffin series, methyl groups strongly decreasing the rate. This is probably due, either to the stabilising influence they have on a peroxide intermediate, or to the production of some more stable intermediate from the easily attacked tertiary carbon atoms.</p> <p>Increase in the surface/volume ratio of the reaction vessel indicates that chain ending on the walls can control the reaction to a large extent. The addition of almost any material which can produce free radicals under the conditions of the experiment, (e.g. aldehyde or peroxide) reduces the length of the induction period and in large amounts raises the rate. Evidently there are many free radicals capable of initiating the chain and if these are produced over a period of time, as in the oxidation of an added aldehyde, their concentration can become large enough to upset the kinetic balance of the reaction and so raise the total rate.</p> <p>The effect of a salt coating on the reaction vessel walls is of great interest. It seems very unlikely that the initiation reaction is interfered with in any my, since both metal wires and loose salt are as efficient as a complete coating. The inhibition must occur through the destruction of an intermediate, possibly a peroxide, since this has a half life time longer than that of the transient free radicals. Salt and metal surfaces might, however, destroy assail radicals of the type HO<sub>2</sub>. Peroxide removal would explain the fact that the efficiency falls off as the temperature is raised towards the high temperature region.</p> <p>The analytical results having shorn that the rate of pressure change is a good measure of the extent of reaction and that the concentration of intermediates is always very small, the cycloparaffin ring must break to give radicals or fragments that are very rapidly converted to final products.</p> <p>From the point of view of the control or prevention of ignition the most significant results may be summarised as follows. Rapid reaction depends upon the slow build up of active intermediates. These are susceptible, in principle, to destruction both by homogeneous additions and by surface action. Homogeneous additions of an effective kind have proved difficult to find. On the other hand a considerable variety of surfaces exert a most marked effect in prolonging the induction period, in some cases almost indefinitely. The action is not improbably connected with the destruction of the intermediate peroxides. The chains by which reaction is propagated almost certainly are initiated on the vessel walls but prevention of this initiation does not seem to be the method of action of the inhibitors. Therefore the control of ignition by salts would be expected to be most effective when the salt is widely distributed thorough the gas. These considerations perhaps indicate the best lines of future work.</p> |
| spellingShingle | Horscroft, R The kinetics of gaseous oxidation reactions |
| title | The kinetics of gaseous oxidation reactions |
| title_full | The kinetics of gaseous oxidation reactions |
| title_fullStr | The kinetics of gaseous oxidation reactions |
| title_full_unstemmed | The kinetics of gaseous oxidation reactions |
| title_short | The kinetics of gaseous oxidation reactions |
| title_sort | kinetics of gaseous oxidation reactions |
| work_keys_str_mv | AT horscroftr thekineticsofgaseousoxidationreactions AT horscroftr kineticsofgaseousoxidationreactions |