628
INDUSTRIAL A N D ENGINEERING CHEMISTRY
VOl. 18, No. 5
Ultra-Violet Spectroscopy of Flames of Motor Fuels‘ Preliminary Experimental Survey By George L. Clark and Walter C. Thee DEPARTYENT OF CHEMICAL
ENGINEERING, MASSACHUSETTS INSTITUTE 09 TECHNOLOGY, CAUBRIDGE, MASS
The ultra-violet spectra are photographed of the detonation, explosion, and combustion flames in an engine of several commercial brands of gasoline, of gasoline containing butyl nitrite, a detonation inducer, tetraethyl lead, a detonation suppressor, commercial addition agents, etc., and of benzene and alcohol. The results represent a preliminary survey of this field of research. Of interest on the physics side of the investigation are t h e wave-length measurements, by comparison with a standard mercury arc spectrum, of the bands due to hydroxyl groups, to CH (or CH2), to CO (Swan, Deslandres, Weston), and possibly to CN, and of the lines of lead from tetraethyl lead, and of copper, aluminum, etc., from the electrodes of the spark plug. Technically important observations in this preliminary survey indicate t h a t : (a) The spectra in this region give excellent promise of successful experimental attack upon the mechanism of the reactions in a n internal combustion engine. (6)The spectra of detonation, explosion, and combustion are different; the first is characterized by a strongly banded structure. ( c ) The mechanism a t various parts of the stroke and the changes with pressure are not yet clearly distinguishable.
( d ) Both tetraethyl lead and butyl nitrite affect t h e spectra; hence the reaction mechanism. The former causes suppression of the strong bands and shortens the ultra-violet, while the latter even accentuates the intensity in the far ultra-violet, thus throwing some Iight on the radiation theory of detonation control. These phenomena are not necessarily tests of the efficiency of various brands of gasoline. (e) The strong lead lines and the absence of lead compound bands show that the metallic lead atoms, or t h e process of decomposition of t h e tetraethyl lead molecule, is the effective agent in t h e reaction propagation, and t h a t combination with the bromine introduced as ethyl bromide takes place subsequent to the oxidation reaction. (f) The three lead lines a t XX4058, 3683, and 3639 appear with so small amounts of the tetraethyl compound in the fuel t h a t they constitute by far the most sensitive test yet devised for the presence of this compound. ( 8 ) The copper lines from the plug spark appear as emission lines in the presence of lead in the vapor, but as absorption lines in its absence. ( h ) The results in general are sufficiently important and clear t o warrant thoroughly quantitative investigation with the spectroscope data correlated with known composition of fuels, pressures, reaction rates, amounts of detonation, etc.
......
HE great theoretical value of the spectroscope as a tron absorption by lead atoms, catalysis of one of two possible scientific instrument in the study of the ultimate oxidation reactions, positive catalysis by lowering ignition nature of matter is generally recognized. All too in- temperature, negative catalysis by decreasing reaction vefrequently, however, has it been called upon to assist in the locity, catalytic poisoning of reaction vessel walls, absorpsolution of very practical problems because of the failure to tion of radiation otherwise accelerating reaction, etc. These apprehend the fundamental scientific background of every theories, together with those of detonation, have been conone of those problems. Midgley and GilkeyZ have called sidered in detail and the results of experimental tests of the attention to the failure of automotive engineers to make use electron theory have been presented by the writers in other of this instrument in the study of the great problem of inpaper^.^ ternal combustion, fundamental knowledge concerning which Suffice it to say that some of the theories are entirely is vital to every practical problem of design, power, mate- erroneous, others are inadequate in that they are inapplirials, or fuel conservation. These writers have presented a cable to knock inducers, and others, such as the radiation brief outline of the science of spectroscopy, and of some of and the preferential route (hydroxylation), are not amenable the information that might be expected from an experimental to direct experimental proof or disproof by ordinary methods. In the belief that there is a great need for more definite spectroscopic study of internal combustion. New impetus to fundamental scientific researches is now experimental evidence bearing particularly upon the progafforded by the discovery of the suppression or induction of ress of chemical reactions involved in normal and abnormal detonation or “knock” by traces of various chemical com- combustion , and that spectroscopy is the most promising pounds in motor fuels. In order to know what chemical of all the fundamental methods of approach, a comprehensubstances have the very valuable property of being “anti- sive program of investigation in the visible, ultra-violet, and knock,” it is necessary to know how they act; in order to infra-red spectral regions has been outlined. The purpose ascertain how they act it is further necessary to know what of this paper is to present the first experimental results on detonation is. It is scarcely necessary to point out that the ultra-violet spectra of the actual flames of familiar motor there is no entirely satisfactory theory of detonation; hence fuels in an operating engine, and thus to open up the general there are no satisfactory explanations of its control. There problem by showing the possibilities of the spectroscopic are eight or more t>heoriesof the action of detonation sup- method in a most difficult field of investigation. The spectra pressors and inducers, many of them particularly applied to have been photographed of the flames of several commercial the case of tetraethyl lead-. g., deposition of colloidal lead on gasolines of different sources and tests (gasolines A, B, C , D points and edges, lead particles as multiple spark plugs, elec- in Figure 1, marketed in this territory), of pure alcohol and benzene, of gasoline to which was added butyl nitrite (0.5 1 Received December 16, 1925.
T
2
J . Soc. Automotive Eng., 10,218 (1922).
9
THISJOURNAL, 17, 1219 (1925).
529
May, 1926
fi
i
lb
e D'
0
......
I .. M c i c u i y */E
2 .. Mercury arc biityl 3 8 (:as
+
. . . . . . nitrite
1260
Yellow
EO0
B l u e white
.I- butyl nitrite 4 butyl nitrite i 9 Cas i-
1280 720 s o n in95
+ butyl nitrite
1025 120n
Ulue white
4 G u s .I- hiityl n i t r i t e
1160 250
Blue r d
+ butyl nitrite
1120 632
Blue ierl
8 01s
t i 9
