Full Text
Chemiluminescence
V. N. Kondrat’ev, Leningrad.
In the several years that have passed since the appearance of the first review on chemiluminescence, printed by A. I. Rabinovich (¹) in volume IV of the present journal, research in the field of chemiluminescence has advanced considerably. The study of the spectra of the glow observed in various reactions has made it possible not only to establish the nature of the carriers of these spectra, but in many cases also the character of those elementary processes that lead to the emission of light. This latter circumstance places chemiluminescence among the most important problems of chemical science, since the determination of elementary processes, of the mechanism of a chemical reaction perceived by us as a phenomenon of macroscopic order, is the chief task of any physical investigation in the field of chemistry. The principal advantage of optical investigation lies in the possibility, by directly studying the spectrum of chemiluminescence, of detecting those intermediate substances through which the reaction proceeds and which escape the view of an observer armed with ordinary chemical methods.
If we turn to the experimental data, we shall see that, as a result of a chemical reaction, along with the reagents, entirely extraneous substances may luminesce—substances that take no part whatever in the reaction. Cases of such sensitized chemiluminescence are very frequent. Besides the glow of a fluorescent dye during oxidation
oxydisilane, observed by Kautsky and Zocher (²); this also includes the glow of Hg in the reaction \(\mathrm{Na}+\mathrm{Cl}_2\) (³), the glow of Na in the reaction \(\mathrm{H}_2+\mathrm{O}_2\)¹), etc.
The existence of sensitized chemiluminescence indicates that the excitation of luminescence in a chemical reaction, at least in many cases, is a secondary process, not directly included in the chain of processes of which the reaction is composed. Energetically, however, chemiluminescence is wholly determined by these latter, primary processes. This is why the study of chemiluminescence often leads to the establishment of the very mechanism of the reaction.
Methodology of the Study of Chemiluminescence
Before proceeding to consider the phenomenon of chemiluminescence and the results to which research in this field leads, let us dwell somewhat on the methodology of this investigation. Since, for establishing the mechanism of a reaction, gaseous reactions are of the greatest interest, we shall deal here exclusively with the methodology of studying chemiluminescence in the case of gaseous reactions, especially since this methodology is the most developed. Most authors who have studied gaseous reactions from the standpoint of chemiluminescence have used an apparatus whose idea belongs to Haber and Zisch (⁴), who, by their classical work, opened new paths along which the subsequent study of chemiluminescence proceeded. The principle of this apparatus is as follows. Into a reaction vessel, usually having the form of a wide glass tube, gaseous reagents are introduced. For this purpose there are side tubes soldered to the reaction vessel, each connected with a reservoir containing the given gas. In the case of reactions between vapors of substances that are solid or liquid at room temperature, these side tubes have the form of a test tube into which the given substance is placed. To ensure the entry—
¹) This observation was made by M. V. Polyakov and the author in the Laboratory of Electron Chemistry of the Physico-Technical Institute.
Chemiluminescence
...the vapors of this substance into the reaction vessel, the branch tube is placed in an electric furnace. The elasticity of the reagent vapors is set by the temperature of the furnace. At the point where the two reacting gases (vapors) meet in the reaction vessel, a reaction zone is established, which can be detected by the glow characteristic of the given reaction and by the deposit of the reaction product—in those cases when the latter is solid or liquid. The reaction vessel, or that part of it where the reaction occurs, is heated externally (by an electric furnace), which makes it possible to establish a definite temperature at which the reaction proceeds. To study the glow accompanying the reaction, the reaction vessel is usually provided with a transparent window, through which the chemiluminescence spectrum is photographed.
Fig. 1.
An instrument of this kind is shown schematically in Fig. 1. Here \(A\) and \(B\) denote the branch tubes1 introducing the reacting gases, \(P\) is the branch tube connecting the reaction vessel with the pump, \(F\) is the window; \(Z\) denotes the position of the reaction zone. For heating the part of the apparatus \(Z\), some authors used the following simple method: the tube was wound externally with nichrome wire; by passing a current of suitable strength through the wire, any desired temperature could be obtained inside the tube (in the case of glass apparatus, up to \(450^\circ\text{C}\)). This method of heating the apparatus is convenient also in that it makes it possible to dispense with the use of a window,
which is in many respects inconvenient (owing to condensation of the reaction products on the cold glass, and also owing to absorption of light by the gas located between \(Z\) and \(F\)), and to photograph the emission spectrum from the side, directly through the wall of that part of the tube which bounds the reaction zone. In addition, with this method of heating the apparatus, the study of the distribution of the emission in the reaction zone is considerably simplified, which is very important for studying both the mechanism and the kinetics of the reaction \((^5,\ ^6)\). To solve the latter problem, the reaction zone is photographed from the side with an ordinary photographic camera and, by photometry of the image, the relative distribution of emission in the zone is found.
The investigation of chemiluminescence, in general not difficult from the methodological standpoint, encounters the greatest difficulties because of the low intensity of the emission in the case of most reactions. Only a comparatively small fraction of reactions are distinguished by a comparatively large yield of light [thus, according to Beutler and Polanyi \((^5)\), the intensity of the yellow emission in the case of the reaction \(Na_2 + Cl_2\) reaches several candles of Hefner], and here it appears possible to study the spectra of chemiluminescence fully. Therefore, in the following description of the phenomena of chemiluminescence we shall dwell only on some of the best-studied cases, referring the reader for a more complete survey of chemiluminescence to that printed in Bulletin of the National Research Council 59 (1927).
Luminescent processes.
We shall begin the consideration of the various cases of chemiluminescence with the simplest one, when the luminous center is a molecule arising as the result of a reaction. In the most elementary form such a case is presented in the formation of a molecule from atoms. From considerations on which there is no need to dwell here, it follows that the formation of any molecule proceeds directly upon collision of the corresponding atoms, provided that one of them is excited—
CHEMILUMINESCENCE
… consisting of atoms, and the energy released as a result of the reaction (the heat of reaction) is emitted in the form of light of one wavelength or another. Experimentally, chemiluminescence of this kind was discovered by Kondrat’ev and Leipunsky \((^7)\) in the case of the formation of molecules of the haloids \(\mathrm{Cl}_2\), \(\mathrm{Br}_2\), \(\mathrm{I}_2\). It was these authors who found that, in the spectrum of the glow observed when vapors of the haloids are heated to temperatures of \(800\text{--}900^\circ\mathrm{C}\), along with the line-banded molecular spectrum there appears a region of continuous spectrum. As Kondrat’ev and Leipunsky subsequently showed, the presence of the continuous spectrum must be connected with a reaction proceeding according to the scheme
\[ \mathrm{X}' + \mathrm{X} = \mathrm{X}_2 + h\nu' \tag{1} \]
(\(\mathrm{X}\) is a haloid atom) between haloid atoms appearing in large quantities at high temperature.
An analogous case is apparently represented by the glow observed during the precipitation of \(\mathrm{NaCl}\), \(\mathrm{KCe}\), \(\mathrm{KBr}\), \(\mathrm{KI}\), and other salts from solution, and explained, according to Weiser \((^8)\), by the recombination of ions into a molecule:
\[ \overset{+}{\mathrm{Me}} + \overset{-}{\mathrm{X}} = \mathrm{MeX} + h\nu . \tag{2} \]
Somewhat more complex is the case of the following reaction
\[ 2\,\mathrm{SnCl}_3 = \mathrm{SnCl}_2 + \mathrm{SnCl}_4 + h\nu, \tag{3} \]
studied by Polanyi and Scheel \((^9)\). Finally, here too we may point to the reactions
\[ \begin{aligned} \mathrm{HgX}_2 + \mathrm{Na} &= \mathrm{HgX} + \mathrm{NaX} + h\nu,\\ \mathrm{CuX}_2 + \mathrm{Na} &= \mathrm{CuX} + \mathrm{NaX} + h\nu. \end{aligned} \tag{4} \]
As Kondrat’ev showed \((^{10,6})\), these reactions, under the known conditions, proceed on a solid surface, and the carriers of the spectrum are the molecules \(\mathrm{HgX}\) and \(\mathrm{CuX}\). All the examples of chemiluminescence cited thus far may be united under the term primary chemiluminescence (Polanyi), since the excitation of the glow and the reaction itself
here are caused by one and the same elementary process. This also includes certain cases of chemiluminescence associated with the so-called triple collision. Here, however, the luminous centers are, generally speaking, extraneous particles. A triple collision consists in the fact that two reacting particles, colliding with one another, undergo a simultaneous collision with a third particle, which takes upon itself the energy released as a result of the reaction. Thus the role of this “third” particle is analogous, for example, to the role of radiation in a reaction of type 1. A triple collision necessarily enters into the reaction mechanism whenever direct formation of the compound in a simple (“double”) collision of the reacting molecules is impossible, i.e., when the excess energy cannot be emitted or when this energy cannot in some way be redistributed within the molecule so that the latter will be “stabilized”1.
The simplest example of a reaction proceeding through a triple collision is the reaction of formation of \(H_2\) from normal \(H\) atoms:
\[ \mathrm{H}+\mathrm{H}+\mathrm{C}=\mathrm{H}_2+\mathrm{C}'. \tag{5} \]
This reaction usually takes place on a solid surface, which plays the role of the “third particle” (\(C\)). In addition to heating of the surface due to the energy released in the reaction (for example, a tungsten wire is thereby heated to incandescence), in this case luminescence of the surface is often observed. Such is the nature of the luminescence of the surface of glass and quartz, as well as of specially introduced fluorescent dyes in active (atomic) hydrogen obtained by an electric discharge \((^{11})\). Let us note that the same luminescence of the surface is observed in hydrogen activated by palladium \((^{12})\). This circumstance is one of the arguments in favor of the supposition that here too we are dealing with atomic hydrogen.
Another, considerably larger, domain of luminous reactions consists of reactions in which the glow is caused by a side process not connected with the reaction as such. This includes all cases of chemiluminescence in which the luminous centers arise upon collisions of already formed, energy-rich reaction products with particles capable of being excited. The mechanism of this kind of excitation of luminescence is essentially no different from the mechanism of impacts of the second kind1. A typical example of this kind of chemiluminescence is the glow observed in the reaction between gaseous sodium and calomel, which proceeds, according to Beutler and Polanyi (⁵), by the following scheme:
\[ \left. \begin{aligned} a)\;& \mathrm{Na}+\mathrm{HgCl}_2=\mathrm{NaCl}+\mathrm{HgCl},\\ b)\;& \mathrm{HgCl}+\mathrm{Na}=\mathrm{NaCl}^{*}+\mathrm{Hg},\\ c)\;& \mathrm{NaCl}^{*}+\mathrm{Na}=\mathrm{NaCl}+\mathrm{Na}'. \end{aligned} \right\} \tag{6} \]
Here, by the processes (a) and (b) the reaction essentially comes to an end, and process (c) is a process entirely extraneous to the reaction. The appearance of excited sodium atoms \((\mathrm{Na}')\) in collisions with the molecules of the reaction product possessing excess energy—\(\mathrm{NaCl}^{*}\)—is connected with the purely accidental circumstance that this excess energy exceeds the excitation energy of the yellow sodium glow. Undoubtedly, there exists an enormous number of reactions proceeding exactly according to the scheme of reaction \([6a, b]\) and not accompanied by luminescence solely because the relation between the excess energy of the reaction product and the energy required for excitation of the particle with which the molecule of the reaction product may collide is not as favorable as it is in the case of the reaction \(\mathrm{Na}+\mathrm{HgCl}_2\). With respect to luminescence, reactions between \(\mathrm{Na}\) and \(\mathrm{HgBr}_2\) (¹⁰), \(\mathrm{HgI}_2\) (¹⁰, ¹³), \(\mathrm{Cl}_2\), \(\mathrm{Br}_2\) (¹³, ¹⁴), \(\mathrm{I}_2\) (⁶, ¹³, ¹⁴, ¹⁵), \(\mathrm{CdCl}_2\) (¹⁰, ¹³), \(\mathrm{CdI}_2\), \(\mathrm{Hg(CN)}_2\), \(\mathrm{PCl}_3\), and \(\mathrm{HCl}\) (¹³), and also reactions between \(\mathrm{K}\) and
HgCl\(_2\) (\(^{10,14}\)), HgBr\(_2\), HgI\(_2\) (\(^{10}\)) and K, Rb and Cs and I\(_2\) (\(^{15}\)). Beutler, Bogdandy and Polanyi (\(^{14}\)), who studied in detail the chemiluminescence spectrum in the reaction of Na with HgCl\(_2\), Cl\(_2\), Br\(_2\) and I\(_2\), found that the principal intensity of the emission here falls to the \(D\)-lines—the first doublet of the principal series of Na. The higher members of the principal series and the lines of the subordinate series, found in the emission spectrum in the case of HgCl\(_2\), Cl\(_2\) and Br\(_2\) together with the \(D\)-lines, are very weak. The intensity of the brightest of these lines amounts to 1.5–0.5% of the intensity of the \(D\)-lines. The excitation energy of most of these lines considerably exceeds the energy liberated in the reaction. The appearance of these lines in the chemiluminescence spectrum was explained by Kondrat'ev (\(^{16}\)) by the special properties of the NaX\(^*\) molecule, which is an electric dipole capable of absorbing energy from the surrounding space in the form of quanta of infrared light. Beutler and Polanyi\(^1\)) connect the appearance of these lines with an entirely different mechanism, namely with the mechanism of two successive collisions occurring between three particles, of which two possess a certain excess energy obtained by them as a result of the reaction, while the third particle (in the present case the Na atom) is the receiver of the energy of the first two particles. The small intensity of the lines that are now of interest to us, in comparison with the intensity of the \(D\)-lines, is explained by the small probability of collisions with two excited particles. The presence of this kind of collision, according to Beutler and Polanyi, must also be credited with the appearance of Hg lines observed in those cases when mercury is present in the reaction zone. For excitation to be effected in this way it is therefore necessary that the particle being excited should collide successively with two excited particles, and that in the interval of time between the two collisions it should not lose the energy acquired in the first collision. Thus this mechanism of excitation reduces to two successive impacts of the second kind.
\(^1\)) Communicated to the author by Prof. M. Polanyi.
Intermediate Substances and the Mechanism of Reactions.
Having completed a brief review of the processes underlying chemiluminescence, in the case of some of the most thoroughly studied reactions we shall turn to the study of the very mechanism of the reactions on the basis of data obtained in the study of chemiluminescence. Establishing the mechanism of reactions amounts to establishing those stages, those elementary processes, of which the reaction is composed. These intermediate elementary processes proceed through compounds which usually, owing to their low stability, cannot be detected by chemical means. Hence the enormous significance of the study of chemiluminescence as a method that gives certain indications of the nature of intermediate substances possessing an extremely short lifetime. The establishment of intermediate substances in many cases is equivalent to establishing the mechanism of the reaction. Let us consider, in several examples, how, on the basis of a detailed study of the spectrum of chemiluminescence, it is possible to construct a probable mechanism of the reaction.
As a first example let us take the reaction between gaseous sodium and chlorine. The establishment of the mechanism of this reaction is due to Beutler and Polanyi \((^{5})\). According to these authors, the reaction \(\mathrm{Na} + \mathrm{Cl}_2\) is composed of the following elementary processes:
\[ \left. \begin{aligned} a)\;& \mathrm{Na} + \mathrm{Cl}_2 = \mathrm{NaCl} + \mathrm{Cl},\\ b)\;& \mathrm{Cl} + \mathrm{Na}_2 = \mathrm{NaCl}^{*} + \mathrm{Na},\\ c)\;& \mathrm{NaCl}^{*} + \mathrm{Na} = \mathrm{NaCl} + \mathrm{Na}'. \end{aligned} \right\} \tag{7} \]
The presence of the first process \((a)\) is directly proved by studying the distribution of the glow and of the NaCl deposit along the reaction zone. Studying the reaction between Na and \(\mathrm{I}_2\), analogous to this reaction, Beutler and Polanyi \((^{13})\) found that, in the case when the vapors of Na and \(\mathrm{I}_2\) enter the reaction vessel as opposing streams (see the footnote on p. 725), two reaction zones are clearly distinguished: a narrow dark zone on the iodine side, distinguished by a high density of the deposit
sodium iodide, and a broad zone adjoining the first on the sodium side, white from the NaI precipitate, but glowing with a bright yellow light (the \(D\)-lines). Fig. 2, taken from the work of Kondrat’ev \((^6)\), may serve as an illustration of what has been said. Here the abscissa gives the distance, in centimeters, along the axis of the apparatus from the point corresponding to the maximum of the yellow Na luminescence (arrow); the ordinate gives the relative distribution of the NaI precipitate. As is seen from the figure, the maximum of the precipitate and the maximum of the luminescence are displaced relative to one another by approximately \(5.5\ \mathrm{cm}\). The region of maximum precipitate corresponds to the first, and the region of maximum luminescence to the second, of the zones observed by Beutler and Polanyi. The presence of these two zones finds its explanation in the reaction mechanism proposed by Beutler and Polanyi, according to which this reaction, like the reaction \(\mathrm{Na} + \mathrm{Cl}_2\), proceeds in the 3rd stage. Stage \(a\) takes place in the first, narrow zone. Calculating from the heats of dissociation of the molecules NaCl (\(93.4\ \mathrm{Cal}\)) and \(\mathrm{Cl}_2\) (\(58.5\ \mathrm{Cal}\)) the heat effect of this reaction (\(34.9\ \mathrm{Cal}\)), it is easy to see that the energy liberated in this reaction is insufficient to excite the luminescence of sodium\(^1\).
Fig. 2.
Consequently, reaction \((a)\) cannot be accompanied by luminescence; accordingly we assign it to the dark first zone. In contrast to reaction \((a)\), in the second stage of reaction \((b)\), which is distinguished by a large heat effect (\(> 60\ \mathrm{Cal}\)), particles rich in energy arise, capable of exciting sodium atoms. This stage of the reaction, and the reaction \((c)\) following it, should naturally be assigned to the bright broad zone. Moreover, the appearance of atomic chlorine as an intermediate product in the reaction between
\(^1\) The energy of the excited \(D\)-lines of Na is \(48\ \mathrm{Cal}\) per gram-atom.
CHEMILUMINESCENCE
with chlorine and sodium was proved more directly by the experiments of Bogdandy and Polanyi \(^{17}\), who detected atomic chlorine chemically by adding water vapor to Na and Cl\(_2\) and observing the appearance of HCl, formed in the reaction of chlorine atoms with H\(_2\) molecules. Beutler and Polanyi also succeeded in showing that reaction \((a)\) occurs at each collision between Na and Cl\(_2\). Beutler and Polanyi arrived at the establishment of process \((b)\) by studying the dependence of the intensity of chemiluminescence (the Na \(D\)-line) on the temperature of the reaction zone. It turned out that the glow weakens with increasing temperature. From this it could be concluded that an increase in temperature destroys those molecules whose presence is necessary for exciting the glow of Na. The supposition that such molecules are Na\(_2\) molecules is supported by the fact that the heat of dissociation of these molecules, calculated from the temperature coefficient of the intensity of the Na glow in this reaction, is close to the heat of dissociation of the Na\(_2\) molecule known from spectroscopic data \(^{1}\). The absence of formation (or the small yield) of NaCl\(^*\) molecules in collisions of Na and Cl atoms leads Beutler and Polanyi to the idea that the process
\[ AB + C = AC + B \ldots \tag{8} \]
in all chemical reactions is the most frequent process, proceeding under the condition of its exothermicity at each collision of the reacting particles.
Analogous studies of the luminescence in the reaction Na \(+\) HgCl\(_2\) lead to the establishment of the mechanism of this reaction, reducible to the sequence of processes represented by scheme 6. The absence of Cl atoms in the reaction zone \([in the presence of hydrogen, HCl was not detected \(^{2}\)]) makes it necessary to exclude the process HgCl\(_2\) \(+\) Na \(=\) Hg \(+\) NaCl \(+\) Cl \(^{3}\), one of two possib-
\(^{1}\) Communicated by M. Polanyi at a meeting of the Council of the Leningrad Physico-Technical Institute.
\(^{2}\) Communicated to the author by Polanyi.
\(^{3}\) This process is included in the reaction mechanism proposed by Kondrat’eva \(^{10}\) and, on the basis of a more detailed investigation of the reaction by Beutler and Polanyi, proved to be erroneous.
... processes initiating the reaction. Thus process 6a proves to be the only possible one. On the other hand, the absence of a dependence of the intensity of luminescence on temperature \((^{5,6})\), in contrast to the reaction \(Na + Cl_2\), leads directly to process 6b, following scheme [8].
Let us consider further the reaction between sodium and \(SnCl_4\), which, under conditions of an excess of \(SnCl_4\), proceeds as far as the reduction of \(SnCl_4\) to \(SnCl_2\). Above we have already given the elementary process which accounts for the luminescence in this reaction. We also noted that this process is an integral part of the reaction. Here we shall give grounds that allow this process to be regarded as more or less probable. The first stage of the reaction is undoubtedly the reaction \(SnCl_4 + Na = SnCl_3 + NaCl\). From the energy of formation of the \(NaCl\) molecule from the atoms \(Na\) and \(Cl\), equal to 93 Cal, it follows that for the detachment of one chlorine atom from the \(SnCl_4\) molecule an energy is required whose upper limit is 93 Cal. Assuming that this energy is approximately equal to 93 Cal, we can, from the heat of the reaction \(SnCl_2 + 2Cl = SnCl_4\), equal to 116 Cal, calculate the energy required for detaching a chlorine atom from the \(SnCl_3\) molecule. Indeed, summing the equalities:
\[ \begin{array}{rcl} +\, SnCl_2 + 2Cl &=& SnCl_4 \qquad +116\\ SnCl_4 &=& SnCl_3 + Cl \qquad -93\\ \hline SnCl_2 + Cl &=& SnCl_3 \qquad +23 \end{array} \]
we find for the required energy the value 23 Cal\(^{1}\).
Hence, for the energy liberated as a result of process [3], from the following equalities:
\[ \begin{array}{rcl} +\, SnCl_3 + Cl &=& SnCl_4 \qquad +93\\ SnCl_3 &=& SnCl_2 + Cl \qquad -23\\ \hline 2SnCl_3 &=& SnCl_4 + SnCl_2 + 70 \end{array} \tag{3} \]
\(^{1}\) From the fact that \(SnCl_4\) and \(SnCl_2\) exist alongside one another, it follows that the fourth chlorine atom in the \(SnCl_4\) molecule must be bound more strongly than the third atom in the \(SnCl_3\) molecule. As we see, this is in agreement with our calculation.
we obtain the value 70 Cal. On the other hand, an investigation of the chemiluminescence spectrum belonging to this reaction shows that the boundary of the spectrum on the side of the short wavelengths corresponds precisely to an energy of 70 Cal. Hence one may conclude that our assumption concerning the energy required for the detachment of a chlorine atom from the molecule \( \mathrm{SnCl}_4 \) is not fanciful. Thus the mechanism of the reaction of interest to us reduces to the following scheme, which was proposed by Polanyi and Schay \((^6)\), who arrived at it as a result of the considerations given above:
\[ \left. \begin{aligned} a)\quad & \mathrm{Na} + \mathrm{SnCl}_4 = \mathrm{NaCl}_3 + \mathrm{SnCl}_3 \\ b)\quad & 2\mathrm{SnCl}_3 = \mathrm{SnCl}_4 + \mathrm{SnCl}_2 + h\nu , \end{aligned} \right\} \tag{9} \]
Here it is also necessary to note that an important argument in favor of the reality of process \([3]—[9b]\) is the continuous character of the chemiluminescence spectrum. This shows that the spectrum observed here is not connected with any definite molecule, but is caused by some chemical process, which in the present case can with a high degree of probability be identified with process \([3]\).
Calculation of Reaction Heats from Chemiluminescence Spectra.
From the last example we have seen how important, in studying the mechanism of reactions by the method of chemiluminescence, is a knowledge of the heats of reactions, in particular, in this example, the heat of the reaction \( \mathrm{SnCl}_3 + \mathrm{Cl} = \mathrm{SnCl}_4 \). Unfortunately, the usual methods of determining reaction heats prove unsuitable precisely in the most interesting cases, when an intermediate compound, not capturable by ordinary chemical means, enters into the reaction equation. An unknown heat of reaction, as occurs in the example just considered, leaves much doubt as to the correctness of the reaction mechanism \([9]\).
However, one may try to approach the question from another side; namely, by investigating the chemiluminescence spectrum, one may try to determine the thermal effect of that process which is the cause of the observed glow,
finite, provided that the mechanism of the reaction is known. It must be said that the study of chemiluminescence spectra from this point of view is in a very rudimentary state. The chief difficulty, which makes investigations of this kind by no means always possible, consists here in our ignorance of the laws obeyed by the processes connected with the exchange of energy between interacting molecules. However, in some of the simplest cases attempts to determine the heats of reactions from chemiluminescence spectra prove successful. Thus, if the reaction mechanism [9] is accepted as corresponding to reality, then the unknown heats of reaction, calculated ultimately from the limit of the chemiluminescence spectrum corresponding to the maximum energy liberated as a result of the reaction, must be regarded as more or less close to the truth. Another, perhaps more successful, example is the determination of the heat of dissociation of hydrogen. As Bonhoeffer (11) and Möller (18) showed, active hydrogen directly excites the \(D\)-lines of Na (excitation energy \(48\,Cal\)) and the OH bands (excitation energy \(93\,Cal\)), whereas the mercury line \(2537^\circ A\) (excitation energy \(112\,Cal\)) always appears together with the bands of the hydride HgH, which is formed only in a heterogeneous reaction on the surface of liquid Hg. Thus the mechanism of excitation of Hg must be different from the mechanism of excitation of Na and OH. If (as is generally accepted) one assumes that the mechanism of excitation of gas molecules by active hydrogen consists in a triple collision (see scheme [5]), then it must be concluded that excitation of the Hg atom by a triple collision is impossible, i.e. that the energy of formation of the molecule \(\mathrm{H}_2\) from H atoms is insufficient for excitation of Hg. Hence for the heat of dissociation of hydrogen one obtains \(112\,Cal\) as the upper limit and \(93\,Cal\) as the lower; as is known, the true value of the heat of dissociation is \(100\,Cal\). Similar investigations of the luminescence caused by active nitrogen, which is presumably atomic nitrogen, give as the lower limit of the heat of dissociation of the \(\mathrm{N}_2\) molecule the number \(230\,Cal\), whereas the true
the value is equal to 262 Cal. We see, therefore, that data on chemiluminescence make it possible not only to establish the order of magnitude of unknown heats of reactions, but also give numerical values of these quantities that are more or less close to the true ones. Undoubtedly, further investigations in this direction will provide abundant material that will supplement our still scanty store of information on the energetics of molecules.
Literature
1) A. Rabinovich, “Advances in the Physical Sciences,” 4, 315, 1925.
2) H. Kautsky und H. Zocher. Z. f. Phys. 9, 267, 1922.
3) H. Fränz und H. Kallmann. Naturwiss. 13, 441, 1925.
4) F. Haber und W. Zisch. Z. f. Phys. 9, 302, 1922.
5) H. Beutler und M. Polanyi. Z. f. Phys. 47, 379, 1928.
6) V. Kondratjew. Z. f. Phys. 48, 310, 1928.
7) V. Kondratjew und A. Leipunsky. Z. f. Phys. 50, 366, 1928.
8) Weiser. Journ. Phys. Chem. 22, 439, 480, 576, 1918.
9) M. Polanyi und G. Schay. Z. f. Phys. 47, 814, 1928.
10) V. Kondratjew. Z. f. Phys. 45, 67, 1927.
11) K. Bonhoeffer. Erg. d. ex. Naturwiss. 6, 201, 1927. See also “Advances in the Physical Sciences,” 8, 1928.
12) M. Polyakoff. Naturwiss. 15, 0, 1927.
13) H. Beutler und M. Polanyi. Naturwiss. 13, 711, 1925.
14) H. Beutler, St. Bogdandy und M. Polanyi. Naturwiss. 14, 164, 1926.
15) K. Ljalikov und A. Terenin. Z. f. Phys. 40, 107, 1926.
16) V. Kondratjew. Nature. 121, 521, 1928.
17) St. Bogdandy und M. Polanyi. Z. f. Elektrochem. 33, 554, 1927.
18) F. Mohler. Phys. Rev. 29, 419, 1927.