Ignition Temperatures

The temperature at which rapid combustion becomes independent of external supplies of heat is termed the ignition temperature. Wheeler defines it as " the lowest temperature to which a mixture of a combustible gas with air or oxygen must be raised in order that the chemical action between the gas and the oxygen can become so rapid as to produce flame." A clear conception of this phenomenon may be obtained by supposing a combustible mixture of gases, such as that of air and the vapour of carbon disulphide, to issue through an orifice into an indifferent atmosphere. If the orifice is surrounded by a ring of platinum wire, which is gradually heated up by a current of electricity, a flame will gradually make its appearance. If, as soon as this is observed, the heating of the wire by the current be discontinued, the flame will disappear; it is, in fact, not self-supporting, but depends on the accessory supply of heat through the electrically heated wire. If now the ring is raised to a higher temperature a brighter flame results, owing to an increased rate of chemical action, and at last we shall reach a point where it is possible to cut off the electric current without causing at the same time the extinction of the flame. This is the true temperature of ignition, the temperature at which the reaction proceeds at a rate just sufficient to overbalance the loss of heat by radiation, conduction, and convection from the burning layer of gases, so that the next layer is put in the same state, and steady combustion proceeds.

The minimum temperature at which the reaction in a combustible mixture of gases becomes self-supporting is termed the sub-ignition temperature. This may not correspond to ordinary ignition. In many cases an ordinary flame, causing more or less complete and rapid combustion, cannot be obtained by heating a gas to its sub-ignition temperature, and in those cases in which such a flame appears it is only produced through the intermediary of a cool flame.

In the case of a few substances the ignition temperature lies at or below that of the atmosphere. Liquid phosphoretted hydrogen, P₂H₄, and certain metallic alkyl derivatives are cases in point. These are spontaneously inflammable. Pure gaseous hydrogen phosphide, PH₃, may be ignited by a tube of boiling water, and carbon bisulphide vapour by a gently heated glass rod, the ignition temperature in this latter case being about 120° C. An interesting case of ignition is afforded by ordinary ethyl ether. Its vapour ignites when mixed with air and allowed to rush into a partly exhausted tube. The conversion of the translational energy of the mixture into heat as the gases enter the tube effecting a rise in temperature sufficient to attain to the ignition point.

In 1816 Davy gave the results of the first systematic attempts to determine the ignition temperatures of the more common combustible gases.

The importance of the subject, particularly in connection with hydrocarbons, will be evident when its bearing upon explosions in coal mines is remembered. Of modern methods of determining ignition temperatures, the following deserve consideration:

A stream of the combustible gases mixed with air is passed through a tube, the temperature of which is raised until the gases inflame.

In the experiments of Meyer and Munch the mixture of combustible gases and air (or oxygen) was passed through a capillary tube to the base of a small glass vessel, in which the ignition was destined to take place, and which was inserted in the bulb of an air thermometer. When the mixture inflamed, the temperature of the gases was calculated from the volume of the gas in the air thermometer.

This method is simple and possesses the advantage of yielding results at atmospheric pressures. But the results are liable to be influenced by the catalytic activity of the walls of the tube. Some of the results obtained in this manner are given in the table.

It is important to remember that these results refer to the ignition temperatures of the gases when in motion. These are not quite the same as when the gases are at rest. Thus, it has been observed that the ignition temperature of detonating gas at 150 mm. pressure falls as the velocity rises to a maximum, after which further increase in the velocity has but little influence. This is well shown by the following data:

Velocity	37	93	130	187	280
Ignition temperature, ° C	601	594	593	592	592

The ignition temperature rises with the pressure (vide infra). The diameter of the tube is without influence between the range 3.6 to 11 mm. With tubes of diameter less than 0.5 mm. no definite ignition temperature has been observed.

Ignition Temperatures as obtained by Method I

Gaseous Mixture.	Ignition Temperature. °C.
Hydrogen-oxygen	550-845
Methane-oxygen	650-730
Ethane-oxygen	606-650
Propane-oxygen	545-548
Ethylene-oxygen	577-650
Acetylene-oxygen	509-515
Propylene-oxygen	497-511
Isobutane-oxygen	545-550
Isobutylene-oxygen	537-548
Coal gas-oxygen	647-649
Carbon monoxide-oxygen	650-730
Hydrogen sulphide-oxygen	315-320

Dixon's apparatus (1903)

A method devised by Dixon in 1903 and employed by Dixon and Coward possesses many advantages over the preceding. The essential features are shown in fig. Air (or oxygen) passed slowly through B and up the porcelain tube, the temperature of which was gradually raised by means of an electric current traversing a spiral of platinum wire wound round the outside of the tube. The combustible gas entering at A ascended the narrow tube and issued at the orifice C just above the thermo-junction T. The temperature was noted at which the escaping combustible gas yielded a visible flame at C. The flowing current not only ensured a constant supply of fresh hot gas, but also the removal of the products of slow combustion. It was found that constant results could be obtained provided the rate of flow of the combustible gas through the orifice C and the diameter of the outer tube D exceeded certain minimum values. Thus, for example, in the case of hydrogen and oxygen, with an outer tube 45 mm. in diameter and an orifice of 1 mm. diameter, a constant ignition temperature was obtained provided the volume of hydrogen escaping through C exceeded 9 c.c. per minute. With a wider tube a more rapid flow of hydrogen was essential. Catalytic action of the walls of the tubes is clearly reduced to a minimum. It will be observed that the temperature of ignition yielded in this apparatus is not quite the same as that obtained by Method I. It is that to which the gases must be heated separately in order to inflame immediately upon contact.

Experiments carried out with hydrogen and oxygen at pressures ranging from 428 mm. to 1460 mm. yielded interesting results, some of which were as follow:

Ignition Temperature of Hydrogen and Oxygen at various pressures

ressure, mm.	Ignition Temperature, ° C.
427	597
548	598
760	592
880	586
960	580
1030	577
1133	570
1456	564

The ignition temperature is seen to rise but slowly as the pressure falls to almost half an atmosphere; but increase of pressure above one atmosphere effects a considerable depression of the ignition temperature. The rate of depression, whilst well marked up to one and a half atmospheres is much slower afterwards, and at high pressures the ignition temperature would probably become fairly constant at about 550° to 560° C.

The main results obtained at atmospheric pressure were as follow:

Ignition Temperatures

Combustible Mixture.	Ignition Temperature. °C.
Hydrogen-oxygen	582-594
Hydrogen-air	582-594
Methane-oxygen	550-700
Carbon monoxide-oxygen (dried over sulphuric acid)	689-695
Methane-air	650-750
Ethane-oxygen	560-630
Carbon monoxide-air (moist)	644-658
Ethane-air	560-630
Cyanogen-oxygen	803-818
Propane-oxygen	490-570
Cyanogen-air	850-862
Ethylene-oxygen	500-519
Ethylene-air	542-547
Acetylene-oxygen	416-440
Hydrogen sulphide-oxygen	220-235
Acetylene-air	406-440
Hydrogen sulphide-air	346-379
Carbon monoxide-oxygen (moist)	637-660
Ammonia-oxygen	700-860

It is interesting to note that the ignition temperature of hydrogen in oxygen was found the same as for hydrogen in air. The same is true generally for carbon monoxide and ethane; but not for methane or ethylene. Cyanogen and hydrogen sulphide ignite at temperatures in oxygen considerably lower than in air. In the latter case, indeed, there is a difference of 140° C.

It is a matter of common knowledge that when a gas is rapidly compressed, a rise in temperature takes place, the relationship between the initial and final temperatures being given by the expression



when T is the absolute temperature and y the ratio of the gaseous specific heats at constant pressure and volume.

It is clear that if a mixture of a combustible gas and air were employed and the compression were sufficiently great, the temperature might rise to the ignition-point and rapid combustion ensue. This was suggested many years ago by Nernst, and carried into effect by Falk 1 in 1906 and by Dixon eight years later. Falk's method, which may be regarded as of a pioneering character, is open to criticism and has led to untrustworthy results.

Dixon and Crofr's apparatus (1914)

The apparatus employed by Dixon and Crofts is shown in fig., and consisted of a steel cylinder, 56 cm. in length, bored with a central cavity, which at the base was enlarged abruptly so that it could be closed with a steel plate, kept in place by means of a powerful screw. An annular washer of lead served to keep the joint gastight, the lead being squeezed well into place with the aid of the screw. A hole was pierced through the side wall A near the bottom of the narrow bore, and fitted with a steel plunger, so that the cavity could be closed during compression or opened in connection with a gas-holder or the outside air when filling or emptying. A cylindrical piston fitted loosely into the explosion chamber, its lower extremity being fitted with a leather washer and a bronze cap, which made a close sliding fit with the cylindrical walls.

The descent, of the cylinder was centred by the steel collar, and hard chrome steel plates, cut with a slot, could be placed on this collar to stop the piston-head at any point in its descent. The cylinder was held by an iron frame, which rested upon a large concrete bed. It was surrounded by a brass water-jacket, not shown in the figure, for regulating the temperature. The compression was effected by allowing a mass of iron, weighing 76 kilograms (2.5 cwt.), to fall from a given height, usually 1.5 metres (5 feet) on to H. Lanoline was employed as lubricant.

The value for γ was taken as 1.40, and the assumption made that there was no loss of heat during the compression of the gases in the cylinder. Actually that was not quite the case, but experiment showed that the compression lasted only about 0-06 second in the case of electrolytic gas, so that the loss of heat would be small. The effect would be to raise the calculated temperature of ignition. The experiments were carried out by the method of trial and error, by regulating the number of steel collars until an explosion just took place. The initial and final volumes were thus known, and these data, coupled with the initial temperature, enabled the ignition temperature to be calculated from the equation given above. The main results obtained are given in the table.

A study was made of the influence of the initial temperature upon the ignition, but variation between the normal temperature of the room and 100° C. made no appreciable difference. Increase of pressure likewise appeared without effect, although reduction to half an atmosphere raised the ignition temperature from 526° to 549° C. Addition of excess of oxygen beyond that required for complete combustion resulted in a gradual depression of the ignition temperature until, when the gases were in the proportions 2H₂ + 32O₂, no explosion would take place.

The ignition temperatures of Hydrogen and Oxygen as determine by Adiabatic compression

Relative Volumes of the Gases.	Initial Temperature, ° C.	Initial Pressure (Atm.).	Ignition Temperature, ° C.
2H2 + O2	Room	1.0	526
	100	1.0	527
2H2 + O2	Room	0.5	549
		1.0	526
		1.5	526
		2.0	527
2H2 + O2	1.0	1.0	526
2H2 + O2 + O2			511
2H2 + O2 + 7O			478
2H2 + O2 + 15O2			472
2H2 + O2 + 31O2			. . .
2H2 + O2	Room	1.0	526
2H2 + O2 + H2			544
2H2 + O2 + 2H2			561
2H2 + O2 + 4H2			602
2H2 + O2 + 8H2			676
2H2 + O2 + 13H2			762
2H2 + O2	Room	1.0	526
2H2 + O2 + N2			537
2H2 + O2 + 2N2			549
2H2 + O2 + 4N2			571
2H2 + O2 + 8N2			615
2H2 + O2 + 14N2			712

Ignition temperatures of mixtures of electrolytic gas and argon

Gaseous Mixture.	Temperature, ° C.
2H2 + O2	520
2H2 + O2 + Ar	532
2H2 + O2 + 2Ar	545
2H2 + O2 + 3Ar	557
2H2 + O2 + 4 Ar	570
2H2 + O2 + 8Ar	622
2H2 + O2 + 12Ar	674

Addition of excess of hydrogen, on the other hand, served to raise the ignition temperature steadily, with practically linear precision, so that the temperature of ignition in the presence of x molecules of hydrogen, within the limits x = 0 and 13, could be calculated from the expression

(2H₂ + O₂ + xH₂) explodes at (526 + 18x) °C.

Nitrogen behaved similarly, the corresponding expression between the limits x = 0 and 14 being

(2H₂ + O₂ + xN₂) explodes at (526 + 11x)° C.

Fiesel's apparatus (1921)

A fourth method, employed by Fiesel, consists in raising the combustible gases separately to a given temperature, allowing them to mix, and noting by means of a delicate membrane or a manometer whether or not a difference in pressure occurs. Variation in pressure indicates chemical change.

The apparatus consisted of an iron globe C (fig.), in which combustion took place. Into this was fixed a thin-walled glass bulb G, attached to a manometer M, and to the oxygen supply. A small piece of metal is hung in the rubber-pressure tubing at B, and can be released at will. C rests on an iron pillar A, and is kept in position by the iron tube D, which is pressed on to it by screwing up nuts E₁, E₂.

The apparatus was placed in an electric furnace and G filled with oxygen. The temperature was now raised, and hydrogen introduced into C. A steady temperature being attained, as registered by two thermo-couples in C (not shown in the figure), the pressure of the oxygen in G was slightly increased, and the metal piece in B allowed to fall and break G, thus causing the oxygen to pour into the hydrogen. If combustion took place, a variation in pressure occurred which was registered either by the manometer or by a delicate membrane attached thereto.

The following results were obtained:

Ignition temperatures of mixtures of oxygen and hydrogen

Gaseous Mixture.	Dry Gases.°C.	Moist Gases.°C.
H2 + O2	407	407 to 417
3H2 + 2O2	397.5	398 to 420
2H2 + O2	401	401 to 425
3H2 + O2	412	436
4H2 + O2	433	479

The minimum ignition temperature was found to occur with 3 volumes of hydrogen to 2 of oxygen, both in the dry and when moist. Further addition of hydrogen raised the ignition temperature, as was observed by Dixon and Crofts, but the results obtained by the latter investigators were very much higher.

With the moist gases the rate of combination suggested a bimolecular reaction, which might proceed through the formation of hydrogen peroxide. For the dry gases the reaction was found to be trimolecular, as is to be expected from the equation:

2H₂ + O₂ = 2H₂O.

The results for acetylene were not altogether satisfactory, the ignition temperature of a mixture of acetylene and air appeared to be about 390° C.

Thornton's apparatus (1919)

Attention has already been directed to the influence exerted by hot, solid surfaces upon gaseous combustion. Mallard and Le Chatelier examined the effect of heated wire gauze upon the combustion of firedamp, and several later investigators have studied the problem of gaseous ignition in contact with hot wires. These researches have been mainly concerned with methane and firedamp. More extensive experiments were carried out by Thornton in 1919, who admitted various combustible mixtures to a small glass vessel A (fig.) of capacity 50 c.c. Thin wire B, soldered to thick copper leads C D, was then rapidly heated by an electric current. It was observed that there was for each diameter and metal a particular current which just caused ignition, provided the temperature rose suddenly and not gradually. The lowest current required for igniting the gases under these conditions was noted, and from this it was easy to calculate the temperature of inflammation. It was found that the phenomena of surface combustion played an important part, and temperatures of ignition were very much lower than those obtained by the methods previously described. Thus, for example, the ignition of hydrogen in air began when the temperature of the wire did not exceed 213° C., nearly 400 degrees lower than the value found by Dixon and Coward, namely, 582° to 594° C. The influence of variation of pressure between 20 and 600 cm. of mercury was negligible, and the proportion of combustible gas exerted in general but little effect upon the igniting current. Clearly, therefore, these results are more comparable with those connected with surface combustion than with the ignition data determined by the methods previously described.

In 1917 McDavid suggested that ignition temperatures might be determined by allowing the mixture of combustible gases and air to inflate soap bubbles and causing these to impinge upon a heated platinum wire. The short time of contact with the wire should reduce to a minimum any surface action such as that detailed above. The results obtained, however, are of uncertain value.