Candle Flame

The cnadle flame

The so-called " wax " of a candle is a solid fuel, with carbon and hydrogen as combustible constituents. When once the candle has been lighted, and equilibrium has established itself, the wax at the foot of the wick melts by the heat radiated from the flame, and ascends the wick by capillary attraction. Arrived in the flame, it vaporises with partial decomposition, yielding a combustible gaseous mass, apparent to the outside observer as a non-luminous inner zone. By inserting a short, narrow glass tube into that portion of the flame, the supply may be tapped, and the vapours ignited at the free end of the tube. As the vapours ascend the cone, slow combustion begins and admixture with a little air takes place. At the apex of this cone combustion begins to be vigorous, for air has now diffused towards it, and the temperature rises towards 1000° C. Above this apex, and forming also a thin mantle all round the lower portions of the cone, is a brightly luminous mantle. In this region partial combustion is taking place and the temperature ranges from 1000° to 1300° C. As the gases flow towards the outer surface of this mantle they meet fresh supplies of air to enable complete combustion to ensue. This takes place in what is termed the non-luminous outer mantle, and it is in this region that the flame is hottest. Here, too, the bent wick, already charred, projects its end, which becomes completely oxidised so that the services of snuffing-tongs are no longer in demand. Just beneath the wick lies a small blue zone which can easily be distinguished but is of relative unimportance. Air rising up the outside of the candle keeps the walls of wax cool and thus prevents guttering.

In 1815 Davy suggested that the Cause of Luminosity of a candle flame is due to the presence of minute particles of carbon at white heat. These particles were believed to be produced by incomplete combustion of the hydrocarbon vapours in the restricted supplies of air available within the flame, the hydrogen of the vapours being " preferentially " oxidised, leaving the carbon to shift for itself. This theory was generally accepted for many years, and it was not until 1867 that a rival theory was projected by Frankland, according to which the luminosity of the flame "is due to radiations from dense but transparent hydrocarbon vapours."

The relative merits and demerits of these theories may most advantageously be considered by reviewing a few of the more important physico-chemical phenomena of flame.

• The deposition of soot upon a cold object introduced into a candle flame is a familiar obstacle to the adoption of this otherwise convenient method of heating small bodies. The deposition occurs only when the object is heated in the luminous zone, the outer non-luminous mantle in general yielding no soot. This was advanced as an argument in favour of the existence of carbon particles in the flame. Whilst this is exactly what might be expected in such circumstances, it would also result upon the decomposition of dense hydrocarbons under like treatment. Hence this experiment alone is in harmony with both theories, and does not enable a distinction to be made.
• The preferential theory of combustion is not now regarded as correct, having given place to the more satisfactory association theory of Bone. But whilst Davy's theory of luminosity thus loses a certain amount of support, it receives support from other directions. Thus it is well known that the higher hydrocarbons are apt to undergo decomposition at high temperatures. In technical practice this is termed "cracking," and it is quite conceivable that analogous reactions might occur in a candle flame.
• When an intense beam of light is projected on to a candle flame the beam is both bent from its original direction and polarised. In other words, the flame behaves as though it were a turbid medium, that is, one containing minute particles, termed the disperse phase, floating about in a continuous phase.
  
  If the diameters of the disperse particles, supposed spherical, lie between the limits of 1 and 100 μμ (μμ = 10^-6 mm.) the particles are of colloid dimensions, and the flame is colloidal. It is termed a stationary, but not a stable dispersoid system by von Weimarn. In a stable system the particles would not change, whereas in a flame they constantly disappear and are as frequently renewed.
  
  Whilst, however, this result clearly indicates the diphasic character of the luminous zone, it does not indicate the nature of the disperse phase. Hence it does not enable a decision to be made between Davy's theory and that of Frankland.
• Similarly, indecisive results are obtained by the spectroscopic study of the luminous zone. A continuous band of colour is observed, and this would result whether the luminous particles were solids or dense hydrocarbon vapours. As is mentioned later, even the flame of hydrogen burning in oxygen under high pressures yields a continuous spectrum, and in this case the possibility of solid particles being present is entirely ruled out.
• The Cause of Luminosity of a flame can be greatly increased by the introduction of solid particles which become incandescent, and the rapid combustion of such substances as give non-volatile solid oxidation products is usually accompanied by brilliant luminosity. A familiar example is the combustion of metallic magnesium. But hydrogen burns in oxygen under pressure with high luminosity, so that solids are not essential to the phenomenon.

From the foregoing it will be evident that a decision between the theories of Davy and Frankland cannot be easily arrived at. Indeed, it is by no means impossible that both theories are correct in so far as they go. The only really certain feature is that the luminous zone is diphasic.

Either one or both of two difficulties may be encountered in determining the amount of dissolved oxygen by chemical methods, even if the modified apparatus of Letts and Blake (Proc. Roy. Dublin Soc., 1899-1920, 9, 454), or of MacArthur (J. Physical Chemistry, 1916, 20, 495) is employed. These a;re (i) closing the apparatus after filling with water and the reacting chemicals without including a bubble of air; and (ii) opening the apparatus and adding the necessary acid to stop further oxidation without admitting some air. The apparatus shown in the Figure, overcomes these difficulties, and is easy to work. The space between the stopper and the 250 cc. mark need not be exactly 2 cc., to arrange which would be a matter of considerable difficulty in view of the character of the stopper. Using Winkler's method, the flask is filled to the 250 cc. mark with water, and approximately 1 cc. of solution added, containing 33 grs, NaOH, and 10 grs. KI per 100 cc. About 2 cc. of solution, containing 40 grams MnCl₂.4H₂O per 100 cc. are now added, sufficient to raise the level above the lower end of the stopper and to fill the capillary tube, when the latter is pushed home. On closing the tap the apparatus is sealed to the entire exclusion of air.

After the reaction has taken place in this flask, some 5 cc. conc. HCl are added to the funnel, the tap opened, and the stopper slowly withdrawn. The acid passes down the capillary and stops further reaction in the flask before air can gain access. A rubber-stopper may now be inserted and, after shaking and standing for five minutes, the liberated iodine is estimated preferably in another vessel, with N₁₀ thio sulphate.