December, 1926
INDUSTRIAE &$B ENGINEERING CHEMISTRY
can be maintained very steady. This is in contrast with the alternating current arc, which is difficult to hold steady and from which the gases come hot. Not only does this obviate the simultaneous ionization of air, hydrocarbon, and catalyst vapor, but the passage of the ionized gas through cooling coils undoubtedly reduces the ionization. The removal from between the primary and secondary ionization chambers (Figure 1) of a 20-cm. (%inch) length of 10-cm. glass tubing reduced for 25 mm. (1 inch) of length a t each end increased the ionization fivefold. It should be mentioned that the presence of X-radiation has practically no effect on most of the atoms over which it passes, but that its effect is concentrated on the particular atoms which it ionizes. The development of the technic of accurate measurement of gaseous ionization by means of the ionization chamber and electrometer suggests new experiments and applications. Of scientific interest are the studies of ionization potential of various gases, and the effects on ionization of changes in temperature and pressure and the presence of a gaseous catalyst. A very important application is the measurement of X-ray intensities for dosage in X-ray therapy. P r e s e n t Status
The fact that many theories have been advanced to explain detonation and the action of antiknock compounds is an indication of the lack of knowledge of these phenomena. Detonation has been considered in detail in another paper in this symposium by Clark and Thee. The theory of “knock” that is subject to least objection is the detonation wave theory, according to which high pressures in the wave front give rise to arbitrary motions producing sound.
Theories of Action of Knock Suppressors Some of the theories of the action of knock suppressors are as follows: 1-According to a French theory, colloidal lead formed in the vapor by decomposition of lead tetraethyl deposits on points, corners, and edges, which may help to accelerate reaction velocity in the flame to the detonation stage. This does not explain why knocking begins immediately if fuel alone is substituted for the mixture containing the suppressor, though the lead and its oxides continue to cling to the walls; nor why particular compounds of lead exhibiting organic valences are required; nor why organic amines, etc., are also effective. 2-The lead particles become incandescent by chemical decomposition of the compound and act as multiple miniature spark plugs producing uniform combustion throughout the mixture. This is subject t o the objections t h a t i t does not explain why Pb(CzH&,CI and Pb(C2H6)2Cllr respectively, are threefourths and one-half as effective as Pb(CzHs)r, nor does i t explain the nature of the decomposition, the action of the amines, and of knock inducers. 3-The lead atom absorbs electrons in the wave front, thus retarding velocity of propagation. This theory has been proved incorrect experimentally in this paper. 4-Knock suppressors catalyze one of two or more possible oxidation reactions-e. g., hydroxylation of hydrocarbonswhich is least likely t o result in detonating velocities. &The simple positive catalyst theory postulates t h a t suppressors lower the ignition temperature to the point a t which the heat of adiabatic compression could cause appreciable but relatively slow chemical action to occur ahead of the flame so t h a t the reaction wave would pass through partly burned air. The objection is t h a t there is no relationship between ignition temperature and tendency to detonate. 6-The simple negative catalyst postulates that the substances as gases decrease the reaction velocity of the burning fuel (Midgley). 7-The theory of poisonings of the metal walls of the combustion chamber which are acting as positive contact catalysts of reaction velocity has the advantage t h a t organic amines are known to be mild poisons and iodine, selenium, lead, etc., pronounced ones. The objections are t h a t mercury and sulfur, which are catalyst poisons, are slight knock inducers, t h a t no
* Schlesinger,
J. SOC.Automoliuc Eng., 16, 441 (1925).
iiig
appreciable lag exists between abwtac@and presence of knock when fuel without suppressor is suddenly introduced, and t h a t detonation occurs as well in glass as in metal vessels. 8-The radiation theory is t h a t the gaseous knock suppressors absorb radiations from the initial flame, which otherwise may activate and accelerate reactions by splitting hydrocarbons into a more reactive condition. This theory is without experimental proof, and will be difficult to test. It would seem inadequate to explain inducers.
Suggestions for Further Work
It is clearly evident that much more experimental work is necessary before a really satisfactory mechanism of the action of knock inducers and suppressors may be devised. Some of the fundamental experiments suggested by the various theories which have been proposed are as follows: (1) Effects of knock inducers and suppressors on speed of propagation of flames photographically registered. (2) Laboratory experiments on reactions of knock suppressors such as lead tetraethyl with hydrocarbons and oxygen, for light bearing upon the nature of organic valences, which must be an important factor in the antiknock reaction. (3). Accurate spectrometric studies of combustion and detonation of various fuels in actual engines and in presence of various knock inducers and suppressors a t particular points in the cycle. This is now being done utilizing an engine with quartz window and a synchronous shutter. (4) Laboratory experiments upon the activation and decomposition of hydrocarbon mixtures by radiation in the spectrum from X-rays t o radiant heat, utilizing radiation from flame itself as analyzed spectrometrically and spectrobolometrically, in the presence and absence of knock inducers and suppressors. ( 5 ) Absorption spectra of tetraethyl lead and related compounds. (6) Adsorption and catalysis studies of fuel mixtures on metal surfaces and possible catalytic poisoning by knock suppressors. (7) Effect upon ignition temperatures of knock inducers and suppressors, using both the secondary discharge and heated surfaces. (8) Relative efficiencies of introduction of knock suppressor in gasoline or in lubricating oil t o test whether existence in gaseous phase is essential or whether the action is a surface catalysis or catalytic poisoning.
Gaseous Explosions 11-Homogeneous and Heterogeneous Reactions Defined and Classified By George G r a n g e r Brown UNIVERSITY OF
MICHIGAN. ANN ARBOR.MICR.
Gaseous reactions m a y be divided into t w o classes, homogeneous reactions wholly confined to a single gas p h a s e a n d heterogeneous reactions o c c u r r i n g at the interface between t w o phases. Homogeneous reactions are c o n t i n u o u s l y homogeneous if all parts of the gas p h a s e are in the same stage of reaction at the same time, progressive homogeneous reactions if the zone of reaction advances c o n t i n u o u s l y through the gas, o r hetero-homogeneous if indirectly modified by another phase. Heterogeneous reactions are gas-gas if the reaction occurs at the interface between t w o gas phases, gas-liquid if the reaction occurs at the interface between a gas phase and a liquid phase, or gas-solid if the reaction occurs at the i n t e r f a c e between a gas phase and a solid phase.
HE properties of gaseous reactions and explosions occurring under different conditions vary in an apparently inconsistent manner. I n a previous paper’ the conflicting evidence concerning the effect of initial temperature upon the rate of rise of pressure of a gaseous explosion was briefly reviewed, and all the apparently conflicting evidence
T 1
Brown, Leslie, and Hunn, THISJ O U R N A L , 17, 397 (1926).
1230
INDUSTRIAL AND ENGINEERING CHEMISTRY
obtained from experimental work with quiet mixtures in a bomb was reconciled by the recognition of the fact that there is a critical initial temperature that gives the maximum rate of rise of pressure in a homogeneous gaseous explosion. I n internal combustion engines conditions are such that an increase in initial temperature usually increases the rate of rise of pressure. There is also evidence that the rate of rise of pressure on ignition may increase as the initial temperature is lowered in cold engines. Throughout this discussion a gaseous explosion means a sudden rise in pressure due to exothermal gaseous reaction, as distinct from a sudden fall in pressure as of an “exploding” steam boiler. Such explosions are of great practical importance, particularly in internal combustion engines. Because of the complicated nature of these reactions and the large number of variables that must be considered, workers in this field find it difficult to meet on a common ground and make many conflicting statements of fact as well as theory. If these reactions can be definitely classified and the properties of the different types studied, order may be brought out of the present chaos. Gaseous reactions are susceptible to a natural classification that brings reactions with the same properties into the same class and eliminates from that class other reactions of distinctly different properties. A careful review of the known properties of gaseous reactions indicates that such reactions are no exception to the major classification generally adopted for all chemical reactions. This classification into homogeneous and heterogeneous reactions has been found to aid greatly in a systematic study of this subject. Exactly where to draw the line between homogeneous and heterogeneous gas reactions is not obvious, and such distinction can be made only upon the basis of some clear-cut, easily defined classification. Homogeneous Reactions
A homogeneous reaction is a reaction wholly confined to a single phase,2 which must be homogeneous before and after the reaction, although not necessarily completely homogeneous while the reaction is taking place. A phase is any homogeneous part of a system, bounded by a surface and mechanically separable from the other parts of the system. “Such a phase is necessarily in one of the three physical states of aggregation, gaseous, liquid, or solid.”2 It does not necessarily follow that a reaction between gases is a homogeneous reaction because a gaseous phase is a homogeneous part of a system. A gaseous reaction is a homogeneous reaction only when the reaction is wholly confined to a single gas phase, and independent of the condition or extent of the surfaces or phases bounding this gas phase. If these surfaces have any direct effect on the reaction, other than to limit the extent of the gas phase, the reaction is not wholly confined to a single phase and cannot be classed as a homogeneous reaction. But the fact that the surfaces limit the extent of the gas phase even in a homogeneous reaction accounts for an indirect effect of the bounding surfaces on a reaction wholly confined to a single gas phase. This type of reaction is discussed as a hetero-homogeneous reaction. Even among those reactions that are wholly confined to a single phase, at least three different types of homogeneous reactions can be recognized. (1) Continuously Homogeneous Reactions
Reactions occurring entirely within a gas phase, all parts of the gas phase in the same stage or degree of completeness of reaction a t the same instant so that the gas phase is com2 Hill in, “A Treatise on Physical Chemistry,” edited by Taylor, pp. 343 and 370. D. Van Nostrand Co.. New York.
ous. T h i s m a y be made clear by consider- A
Vol. 17, No. 12
Decernber, 1925
I N D U S T R I A L A N D ENCZiVEERINC C H E M I S T R Y
sentially a continuous succession of different continuously homogeneous reactions. For these reasons a reaction ocourring entirely within a gas phase, the zone of reaction advancing through the gas in the manner described, is considered a progressive homogeneous reaction, or homogeneous reaction of the second class. (3) Helero-homogeneous Reaction or Compound Progressive Homogeneous Reaction
A reaction wholly confined to a single gaseous phase must be a homogeneous reaction by definition. But if tho reaction is greatly influenced and its properties modified hy the surface bounding the gas phase, although the reaction does not occur a t these surfaces, the reaction is not a simple homo.. geneous react.ion of either class (1) or (2), hut a complex homogeneous reaction defined as a “hetero-liomogeneous reaction” hecause it is wirolly confined to a single gas phase but partially determined or influenced by the surfaces of another pliase (or phases) hounding the gas phase. Mallard and LeChateliers photugaphed the flame passing through an explosive mixture along a horizontal tube, on photographic paper moving vertically. These photographs show three distinct stages of flamc propagation in horn+ geneous gaseous explosions in a closed tube, and only two stages in those explosions fired near the closed end of a tube open a t tho far end. These stages are as follows: l---Initial accelerating continuous propagatioii (a progressive homogeneousreaction), 2-Vibratory movement which either cxtinguishes the flame or develops into the third stage (a hctero-homogeneousreaction). Soir-This stage i s nrver observed in inixtitirs fired near the closed rrid of a tube oprri r i tire far end; t l m flame is then uniformly accelerated until the detonation w i v e i s developed.
wave. A very rapid, intensely luminous high is indcpendent of the material or diameter of the combustion tube, but determined by the conditions o i the (A progressive homogencous reaction.) 3-Detonation
pressure wave propagated at a constant velocity that
The fact that the vibratory stage in a homogeneous gaseous explosioii is eyident only whcn the far end of the tube is closed, is direct evidence that this vibratory stage is dependent upon the surface bounding the gas phase at the far end. Dixon’ has demonstrated that a pressure wave resembling a sound wave is initiated when an explosive mixture is ignited and that this pressure wave reflected from the far end of the tube has a pronouiiced effect upon the combustion and flame propagation in the explosive mixture. I n a t.uhe closed at both ends and containing an explosive mixture of 2He O2ignited a t the center, these pressure waves are reflected back and forth throughout the length of the tube, causing the vibratory movement of the flame as shown in the photograph of Figure 2.* Inaddition to the vibratory movement, Dixon’s work shows that these pressure waves have an accelerathig or intensifying effect upon the combustion because of the greater concentration of the gas in the pressure wave. This can be seen clearly in the photograph. Not only do these pressure waves in an explosive mixture modify or intensify the combustion taking place, but the pressure waved may ignite the unburned gas ahead of the flame by adiabatic compression in much the same way that a detonator cap sets off the main charge of a solid explosive. This “auto-ignition” of tile explosive mixture ahead of the Bame occurs probably where the pressure wave is reflected from some solid surface or where two pressure waves meet,
+
* A n n . mines, IS1 4, 274 (1883); Ann. ihim. phys.. L51 IS, 289 (1883); Cornpi. rend., 91, 825 (1880); 95, 145 (1881); 96. 566 (1882). Berthdot, D i d . , 93, 18 (1881). 7 Trons. ROY.Soc. London. PQQA,319 (1903). a J . SOL.Aulomoiivc Eng., 9 , 237 (1921).
1231
because then the intensity of the adiabatic compression is the sum of the two compressions in the two waves or in the wave and the reflected wave. Dixon* has published a photograph showing the movements of the flame in a mixture of hydrogen with three volumes of oxygen tired by a spark near the right side of the tube while the gases were being compressed by the sudden thrust of a piston from the left. At the moment the mark was passed the gas-had been c o m p r e s s e d t o one-quarter of its origi n a l volume. T h e pressure wave started by the initiation of the explosion reached the advancing face of the piston before the flame, ignited the gas at the face of the piston, and on reflection checked t h e a d v a n c e of t h e flame, initiated by the spark. I n this case the reaction or explosion is entirely confined to the gas pliase and is theref o r e a homogeneous I reaction: but because t h e d o u b l e ignition FlWre L---Phorcrgraph snd Inferpre(auto-ignition) delrei- tation a i Movemenfa of Gas Pssrfirtes durins .Combustion oped by the pressure The eiploiion occurred iu s horizontal wave was dependent tube closed at both ends. The mixture (ZHI + ih) was ignited i n the center of the upon this p r e s s u r e tube and the photographic film WOQ moved downward amo~spthe hoiizontd wave being reflected by vertically window in the tube. the surface bounding the gas phase, the reaction is not entirely independent of the bounding surfaces and is defined as a hetero-homogeneous reaction. Further evidence”‘O of these types of reactions is so complete that this classification of homogeneous gaseous reactions is presented without further qualifications. Heterogeneous Reactions
A heterogeneous reaction is a reaction occurring at the interface between two phases. Any reaction dependent upon two substances not contained in a single homogeneous phase coming into contact can occiir only at the interface between the tw$phases and is a heterogeneous reaction. I n general, heterogeneous reactions are classified according to the phases involved. (1) Gas-Gas Helerogeneous Reaction
A gas-gas reaction is a reaction occurring a t the interface of two gaseous phases. It has generally been assumed that but one gas phase can he present if no restraining containers are used to separate the gases. This assumption, fully justified for systems at equilibrium, is not always true under actual conditions, which may be far from equilibrium. If city gas is allowed to escape from a gas cock into the air and the gas is ignited, the combustion, which can occur Poley and Souder, Phyr. Rcv.,36,373 (1912). pxesent photographnactually shoring t h e round or pressure r a v e sdvancing from the spark through the explosive mixture ahead o i the flame and hat gases, and the reaection of there pressure waves from the ~ r l l 0% s the contrioer. Woodbury, Lewis, and Canby, J . Soc. Auioomoiiue En& 8. 209 (1921); Brsdshaw, Proc. Roy. SOL.London, 7911,236 (19071; Ellis and Wheeler. Furl, 4, 356 (1923).
1232
INDUSTRIAL A N D ENGINEERING C H E M I S T R Y
only when gas and air come into contact, obeys all the laws of a heterogeneous reaction. I n this case the velocity of the reaction depends upon the rate of diffusion of the air and gas to the interface and upon the rate of diffusion of the combustion products away from the interface, as well as upon the rate of chemical reaction in the interface. If the gas is emitted with a higher velocity, introducing turbulence and eddy currents into the flow stream, the rate of combustion is increased because the rates of mixing are increased. This principle is recognized and widely used by combustion engineers. A gas-gas heterogeneous reaction may be of importance in gaseous explosions. I n internal combustion engines this type of reaction might be expected in engines using a stratified charge, and in the fuel injection engines such as those working on the Diesel cycle, in which the air used for combustion is compressed to a high pressure and temperature in the engine cylinder and the fuel subsequently injected, vaporized, and burned during the power stroke. A high degree of turbulence, however, causes such intimate mixing that the gasgas heterogeneous reaction rapidly approaches homogeneous reaction as the rates of mixing exceed the rate of a homogeneous chemical reaction. (2) Gas-Liquid Heterogeneous Reaction
A gas-liquid reaction is a reaction occurring a t the interface of gaseous and liquid phases. I n general, the same statement may apply to gas-liquid reactions as was made concerning gas-gas heterogeneous reactions. I n fact, evidence indicates t’hat liquids are vaporized before oxidation takes place in most combustion reactions. Whether or not this is always true is probably of no importance, as in either case the reaction is a surface or heterogeneous reaction and therefore governed by the same laws. It is well-known that oil fuel may be burned more quickly and completely if finely atomized and injected with turbulent flow into the furnace or engine than if introduced as a viscous stream. If the liquid is uniformly distributed throughout the mixture in the form of a fine mist, the explosive reaction takes place in a mixture that may be considered essentially homogeneous and the reaction approaches a homogeneous reaction. This statement is directly supported by the work of Haber and Wolff,ll who found that the properties of an explosion occurring in a fuel-mist air mixture are similar to the properties of a progressive homogeneous gaseous explosion. (3) Gas-Solid Heterogeneous Reaction A gas-solid reaction is a reaction occurring a t the interface of gaseous and solid phases. I n addition to reactions between gases and solids as dust explosions and explosions in a coal calorimeter bomb, the catalytic action of solid surfaces on explosions of gaseous mixtures is extremely important.12 Many reactions between gases are greatly catalyzed by solid surfaces, particularly when the surfaces are heated to high temperatures. ??le socalled “surface combustion” furnace is the extreme application of this principle to industrial furnaces. Many investigators have emphasized the important effect of hot surfaces on the combustion of explosive mixtures in internal combustion engines. I n general, the engine knock in a high-compression, multiple-cylinder engine may be largely stopped by properly cooling the suspected hightemperature areas.13 As the exhaust valve is .probably the 2. angeu. Chem., 36, 373 (1923). Sokal, J . Sac. Chem. I n d . , 43, 2831‘ (1924). 18 Holloway, Huebotter, and Young, J . SOC.Automofive Eng., 12, 111 (1923); Young and Holloway, Ibid., 14, 315 (1924); 16, 255 (1924); Horning, J . Sac. Automobile Eng., 14, 142 (1924). 11 12
Vol. 17. No. 12
most highly heated area within the combustion chamber, proper cooling of the exhaust valve or its complet,e elimination makes more difference than any other surface factor.l% Catalysis
Midgley and Boyd15 have shown that catalysts mixed with the liquid fuel used in internal combustion engines have very important effects on the explosion. These catalysts become intimately mixed with the fuel and uniformly distributed throughout the explosive mixture as vapor or at least as a fine mist. I n this condition as a part of the homogeneous mixture these catalysts influence all homogeneous reactions as well as heterogeneous reactions involving the explosive mixture. Surface catalysis of gaseous reactions makes the reaction heterogeneous. For this reason all gaseous reactions catalyzed by surfaces are included under heterogeneous reactions. As a large number of gaseous reactions are subject to contact catalysis, catalytic heterogeneous gaseous explosions are veryimportant, particularly in internal combustion engines. 14
15
Abell, J . SOL.Automotive Eng., 13, 301 (1923). THISJOURNAL, 14, 894 (1922).
Length of Visible Flame and Length of Flame Travel in Combustion of Powdered Coal By Henry Kreisinger COMBUSTION ENGINEERING CORP., NEWY o = , N. Y.
T
HE length of flame in the combustion of powdered coal’ is generally understood to be the length of the visible flame in the furnace. The visible flame is produced by the combustion of the volatile matter of coal; that is, it is the combustion of the gaseous combustible in the coal. T h e combustion of the gaseous combustible is made visible by the presence of small particles of incandescent carbon formed. by the breaking down of heavy hydrocarbons. Visible Flame
The length of the visible flame depends on the percentage. of the volatile matter in the coal, the composition of thevolatile matter, the amount of air supplied with the coal, and the rapidity of mixing of the coal and the air. Other factors being the same, coal with a high percentage. of volatile matter will give a longer visible flame than coals having a low percentage of volatile matter, and to the ordinary observer it appears that high-volatile coals require larger furnaces and longer time for complete combustion than coals having a low percentage of volatile matter. I n otherwords, high-volatile coals seem to be more difficult to burn in pulverized form than low-volatile coals. Thus, Illinois coal when burned in powdered form will give a long, dense. flame, which may be somewhat smoky. Pocahontas and Kern River coals will give a bright, intensely hot flame, which will appear considerably shorter than the flame from theIllinois coal. Texas and Korth Dakota coals will also give a long flame, which, however, may not be so long as the flame from the Illinois coal. The flame will be bright yellow and ordinarily will not be smoky. If a sample of the fine dust particles carried by the furnace gases is collected a t the outlet of the gases from the furnace,. the dust from the low-volatile matter coal will contain a higher percentage of unburned carbon than the dust from the high-
