Spectral Study of Photochemical Reactions Occurring Under the Action of Intense Light Fluxes

A. V. Karyakin

In modern chemistry, the mechanism of many reactions can be explained only by taking into account the formation of free atoms and radicals arising from the reagents. Their role in the kinetics of reactions is exceptionally great. The introduction of these intermediate particles into the scheme of successive stages of reactions makes it possible to explain the formation of the final products from the initial ones.

A free radical is a group of atoms split off from a stable chemical compound and possessing an unsaturated valence, i.e., one unpaired electron. Owing to this, free radicals are characterized by addition reactions.

Free radicals and atoms in the gaseous state are usually obtained in the thermal or photochemical decomposition of substances, as well as in an electric discharge in a gas[^1].

There are many methods for detecting them. Some organic radicals can be detected directly by the deposition of metallic sublimated mirrors[^1],[^2]. Thus, for example, free alkyl radicals react with a layer of lead deposited in vacuum. Atomic hydrogen, however, does not react with metallic lead, since lead hydride does not exist, but hydrogen atoms are capable of destroying antimony layers as a result of the formation of antimony hydride. From the rate of disappearance of the mirrorlike metallic deposit one can calculate their mean lifetime under the given conditions. For methyl free radicals, the mean lifetime in the gaseous state is \(8.4 \cdot 10^{-3}\) sec.

Free radicals and atoms have also been detected with the aid of other selective indicators:

a) metal oxides that change their color when acted upon by free atoms (thus, for example, molybdenum trioxide turns blue under the action of atomic hydrogen[^3]);

b) certain luminescent substances that undergo quenching of luminescence under the action of particular free atoms and radicals[^4].

In addition to these selective indicators, the manometric method is used for studying reactions of free atoms.

Thus, free oxygen atoms produced under the action of short ultraviolet rays are detected by their reaction with other gases (hydrogen or carbon monoxide) through the drop in the total pressure of oxygen[^5].

Spectral methods are applied with great success for detecting free atoms and radicals possessing characteristic spectral absorption or emission of light. During the dissociation of molecules under the action of an electric charge or of light and thermal energy, radicals and atoms possessing excess energy are formed in some cases. They give up this excess energy in the form of characteristic radiation, inherent only to the given free atom or radical. This is the most direct method for determining free atoms and radicals, since it makes it possible not only to detect them at the moment of their formation, but also to follow their subsequent fate. Thus, for example, the OH radical was detected by its glow in an electric discharge in water vapor, and also in a hydrogen flame not only in emission, but even in absorption[^6].

Studies of the emission spectra of hydrocarbon flames proved the presence in them of stable radicals OH, CH, C₂, NH, etc., which, owing to their high chemical activity, have a very short lifetime[^7].

Spectral investigations of the glow of vapors of halide salts of metals of almost all groups of the periodic system and of organometallic compounds under illumination by short ultraviolet radiation showed that characteristic emissions of free atoms and radicals take place[^8]. Similar studies were carried out in the photochemical decomposition of molecules of water, alcohols, organic acids, ammonia, and other gaseous compounds under the action of Schumann ultraviolet radiation. In this case free radicals OH, NH, CN, etc. were detected[^9].

However, until recently the principal method for detecting the existence of free radicals and atoms and their participation in reactions has been the study of their kinetics.

R. Norrish and G. Porter have recently developed a new direction in the field of experimental detection of free radicals in photochemical reactions. For this purpose the reactions were induced by the action of a powerful flash of a pulsed lamp, which gives a short-duration luminous flux of high intensity, equal to \(10^{21}\) quanta within \(10^{-3}\) sec. In this case it becomes possible to obtain sufficiently large con-

centrations (from several millimeters to tens of millimeters of mercury) of intermediate reaction products (radicals and atoms in the free state\(^{10,11}\)), and not only to detect their presence spectrally from absorption, but also to study spectrally the reactions between them. The basic scheme of their apparatus is shown in Fig. 1.

The quartz vessel \(A\), in which the reaction takes place, is 1 m long and 20 to 50 mm in diameter; it is connected to a vacuum installation and filled with the initial gaseous compound at a pressure of several millimeters of mercury.

The flash lamp \(Б\), which induces photochemical reactions, in particular decomposition reactions, is arranged parallel to the vessel and is enclosed together with it in an aluminum tube, the inner

Fig. 1. Schematic of the apparatus.

surface of which is coated with magnesium oxide and serves as a reflector. The flash lamp, which is a quartz or glass tube with tungsten electrodes sealed into its ends, produces a powerful pulsed spark—a flash in an inert gas—by discharge of high-voltage capacitors with an energy of up to 10,000 joules.

Absorption spectra are obtained with the aid of another, less powerful flash lamp \(В\), whose light is focused on the vessel and, after passing through it, falls on the slit of the spectrograph \(Г\), with subsequent recording of the absorption spectrum on a photographic plate.

The flash lamp that induces the photochemical reactions and the flash lamp used to obtain the absorption spectra are not ignited simultaneously; the latter is ignited by means of a time relay with some delay. The durations of the flashes of the flash lamps and the intervals between them are recorded by a photocell connected to an oscilloscope. The delay time between ignition of the flash lamps is varied from zero to several seconds and is achieved by mechanical or electrical methods. This makes it possible to determine the lifetimes of free atoms and radicals. The use of photomultipliers for recording absorption spectra did not yield satisfactory results.

Depending on the nature of the photochemical reactions being studied, flash lamps with different flash energies and durations are used. Table 1 gives representative parameters of flash lamps.

Table I

The pulse lamp used by Norrish and Porter for studying absorption spectra had a flash energy of 700 J with a duration of \(5 \cdot 10^{-4}\) sec (Fig. 2, A).

Fig. 2. Pulse lamps used for obtaining absorption spectra.

The pulse lamp used by Herzberg and Ramsay for the same purposes is shown in Fig. 2, B and has an output of 80 J with a flash duration of \(3 \cdot 10^{-4}\) sec.

The distribution of energy by wavelengths for pulse lamps with an output of up to 500 J corresponds to the energy distribution of a black body at \(t = 6500^\circ\)C. The intensity of the continuous spectrum in the ultraviolet region in a lamp 1 m long does not depend on the wavelength in the range 2600–4400 Å, and also does not depend on the gas pressure (within 40–150 mm Hg) or on the flash energy (in the interval from 800 to 4000 J). At flash energies below 500 J and gas pressures below 40 mm Hg, the continuous spectrum decreases in intensity and is replaced by the line spectrum of the inert gas. The total radiation in the ultraviolet region for lamps with an output of up to 500 J does not depend on the gas pressure down to 50 mm Hg, but below the indicated pressure the radiation falls. For flash energies greater than 80 J, the intensity of the radiat...

tion in the ultraviolet region is represented by the equation

\[ I_n = kC\left(V^2 - V_0^2\right)\ \mathrm{J}, \]

where \(k = 5.14 \cdot 10^{-3}\), \(V_0 = 1420\ \mathrm{V}\), \(C\) is the capacitance in farads, and \(V\) is the applied voltage\(^{16}\).

The energy of radiation in the ultraviolet region can be increased by adding mercury to the inert gas, but the light output is then not stable.

Pulsed lamps with a quartz capillary, having a flash energy of \(80\ \mathrm{J}\), give an almost pure continuous spectrum in the range \(2000\text{--}9000\ \text{\AA}\). In the continuous spectrum there are almost no emission lines, but absorption bands of silicon are observed\(^{15}\).

For measuring photochemically active useful radiation, uranyl oxalate actinometry is used, whose quantum yield remains constant for intensities 1000 times greater than those used previously and 100,000 times greater than the intensities of ordinary mercury-quartz lamps\(^{17}\). From the percentage decomposition of the uranyl oxalate system, which is directly proportional to the intensity of the incident light, the absolute intensity of the pulsed lamp was calculated in the interval from 2000 to 4900 \(\text{\AA}\), and was found to be \(10^{21}\) quanta over \(10^{-3}\ \mathrm{s}\).

Such powerful radiation over one thousandth of a second causes almost complete photochemical decomposition of substances in the gaseous state. This gives high concentrations of atoms and free radicals: Cl, Br, ClO, HS, CS, CH\(_2\), CH\(_3\), CH\(_3\)CO, NH, NH\(_2\), OH, C\(_2\), HCO, CH, etc.

Photochemical decomposition and subsequent reactions occurring over milliseconds are accompanied by the release of a large amount of heat. This heat consists of the excess energy of the dissociation products, the heat released when the excess electronic energy of the dissociation products is lowered, and also the heat released in reactions between intermediate products formed as a result of photodecomposition. It is very difficult to calculate precisely the amount of heat released (as well as the temperature). However, it may be assumed that the amount of heat released is approximately equal to the amount of heat produced by combustion of the given substance, and that the temperature of the whole system reaches \(\sim 1000\text{--}3000^\circ\mathrm{C}\).

The abundantly released heat strongly affects the course of the photochemical reaction; however, it is difficult to establish the nature of this influence on the reaction mechanism. In order to eliminate it, an artificial method of cooling is used, consisting in introducing, into the vessel in which the reaction is to occur, in addition to the substance under investigation, an inert gas at a pressure up to \(1\ \mathrm{atm}\). By this method it is possible to almost completely eliminate the thermal influence on photochemical reactions.

I. APPLICATION OF FLASH LAMPS TO THE STUDY OF CERTAIN REACTIONS

Decomposition of chlorine

The photochemical decomposition of the chlorine molecule into atoms with subsequent recombination is illustrated by photographs of absorption spectra in the ultraviolet region \(^{10, 11, 18, 19}\), shown in Fig. 3. The absorption spectrum of the chlorine molecule disappears immediately after the flash of the flash lamp and reappears to half its intensity after \(15 \cdot 10^{-3}\) sec. If the photochemical decomposition of chlorine occurs in the presence of oxygen, then the chlorine monoxide radical ClO is formed, having an absorption spectrum different from the absorption spectra of the known chlorine oxides.

Fig. 3. Recombination of chlorine atoms: a) immediately after the flash of the flash lamp; b), c), d), e) after 12, 17, 60 msec and 60 sec, respectively.

A photograph of its absorption spectrum is given in Fig. 4. The formation of ClO proceeds through an intermediate chlorine oxide according to the following reaction:

\[ \mathrm{Cl} + \mathrm{O}_2 \to \mathrm{O} - \mathrm{Cl} - \mathrm{O}, \]

\[ \mathrm{Cl} + \mathrm{O} - \mathrm{Cl} - \mathrm{O} \to 2\mathrm{ClO}, \]

and at high temperature:

\[ \mathrm{O} - \mathrm{Cl} - \mathrm{O} + \mathrm{Cl}_2 \to \mathrm{Cl}_2\mathrm{O}_2 + \mathrm{Cl} \to 2\mathrm{ClO} + \mathrm{Cl}. \]

The lifetime of the radical formed is \(4 \cdot 10^{-3}\) sec; the subsequent decomposition proceeds by the reaction: \(2\mathrm{ClO} \to \mathrm{Cl}_2 + \mathrm{O}_2\) and depends neither on the pressure of the radical itself nor on the presence of a foreign gas. If hydrogen is added to a mixture of chlorine with oxygen, or if oxygen is added to a mixture of chlorine with hydrogen, then the chain carriers—Cl and H atoms—disappear, with interruption of the chain reactions. It is assumed that the disappearance of atoms occurs through

formation of \(\mathrm{HO}_2\) by the reaction: \(\mathrm{H}+\mathrm{O}_2=\mathrm{HO}_2\), but it was not possible to detect the \(\mathrm{HO}_2\) radical spectrally in the region \(4750\)—\(2200\ \text{\AA}\). This is evidently due to the fact that the indicated radical

Fig. 4. Absorption spectrum of the ClO radical.

is destroyed according to the following scheme: \(\mathrm{HO}_2+\mathrm{Cl}=\mathrm{HCl}+\mathrm{O}_2\), and has no characteristic spectrum in this spectral region.

Decomposition of diacetyl

Undecomposed diacetyl\(^{11}\) gives a characteristic absorption spectrum in the region \(4000\)—\(4500\ \text{\AA}\) and shorter than \(3000\ \text{\AA}\), as is shown in Fig. 5, \(a\). In addition, Fig. 5 shows the entire process of decomposition of diacetyl, which begins during the flash of the

Fig. 5. Photochemical decomposition of diacetyl: \(a\)) absorption spectrum before the flash of the pulsed lamp; \(b\), \(v\), \(g\)) after 1.2; 2.0; 3.3 msec and \(d\)) 2 min after the flash.

pulsed lamp (Fig. 5, \(b\)) and continues after the flash for \(2\cdot 10^{-3}\) sec (Fig. 5, \(v\)), which is clearly seen from the decrease in absorption in the long-wavelength region and the increase in the short-wavelength region of the spectrum. After \(3.3\cdot 10^{-3}\) sec (Fig. 5, \(g\)) a weakening of absorption is observed in the short-wavelength region, which can

can be explained by the decomposition of the acetyl radical according to the reaction:

\[ \mathrm{CH_3-\dot{C}=O \rightarrow \dot{CH}_3 + CO,} \]

\[ \mathrm{2\dot{CH}_3 \rightarrow C_2H_6.} \]

However, 2 min. after the beginning of decomposition, recombination of a certain amount of acetyl radicals into diacetyl begins, as indicated by the appearance of absorption in the long-wavelength region of the spectrum (Fig. 5,d).

Decomposition of carbon monosulfide

In the photochemical decomposition of carbon monosulfide under the action of radiation from a flash lamp, an absorption spectrum of radicals was obtained (bands at 2445, 2507, and 2575 Å)\(^{11,18}\), previously known only in emission\(^{20}\). At the same time, the spectrum of \(\mathrm{S_2}\) is observed. The CS radical has a long lifetime and, in stability, resembles carbon monoxide (the half-life of the CS radical is 2–3 minutes).

Decomposition of ammonia

Illumination by a flash lamp of ammonia at 10 mm Hg gave an absorption spectrum in the region 5700–6900 Å, in which about 50 lines were observed. This spectrum was assigned to the radical

Fig. 6. Spectrum of the free radical \(\mathrm{NH_2}\): a) in absorption, b) in emission.

\(\mathrm{NH_2}\)\(^{14}\), because it coincided exactly with the emission spectrum of this radical, which had been obtained both in an ammonia–oxygen flame and in the experiments of A. N. Terenin and G. G. Neuymin during illumination of ammonia with Schumann ultraviolet radiation\(^{9}\).

Photographs of the absorption and emission spectra of the \(\mathrm{NH_2}\) radical are shown in Fig. 6.

Decomposition of Ketene

When ketene vapors were illuminated with a flash lamp, the aim was to obtain the absorption spectrum of the radical $\mathrm{CH}_2^{18,21}$. However, the attempt was unsuccessful, apparently because the lifetime of the $\mathrm{CH}_2$ radical is comparable with the duration of the lamp flash.

Illumination of ketene vapors mixed with inert gases by a flash lamp of moderate intensity causes photodecomposition, as a result of which carbon monoxide and ethylene are formed. The mechanism of this decomposition may be represented in the form of the equations:

\[ \mathrm{CH_2CO} + h\nu \to \mathrm{CH_2} + \mathrm{CO}, \]

\[ 2\mathrm{CH_2} \to \mathrm{C_2H_4}, \]

\[ \mathrm{CH_2} + \mathrm{CH_2CO} \to \mathrm{C_2H_4} + \mathrm{CO}. \]

If, however, the light intensity of the flash lamp is increased or the pressure of the inert gas present is reduced, then the photochemical decomposition of ketene gives, in large quantity, hydrogen and acetylene, as well as carbon; in this case the quantum yield of the reaction reaches 6, and the percentage of decomposed ketene becomes higher than 80%. In this case, the action of heat is added to the photochemical decomposition; it also creates free radicals in a concentration approaching the concentration of ketene molecules. As a result, the probability of radicals reacting with one another proves greater than that with ketene molecules. The quantity of heat released in this reaction may be comparable with the quantity of heat released in a ketene flame and is approximately equal to 100 kcal/g mole. The mechanism of the reaction is then as follows:

\[ \mathrm{CH_2CO} + h\nu \to \mathrm{CH_2} + \mathrm{CO}, \]

\[ 2\mathrm{CH_2} \to \mathrm{CH_3} + \mathrm{CH}, \]

\[ \mathrm{CH_3} + \mathrm{CH_2CO} \to \mathrm{C_2H_5} + \mathrm{CO}, \]

\[ \mathrm{CH} + \mathrm{CH_2CO} \to \mathrm{C_2H_3} + \mathrm{CO}, \]

\[ \mathrm{C_2H_3} \to \mathrm{C_2H_2} + \mathrm{H_2}, \]

\[ \mathrm{C_2H_5} \to \mathrm{C_2H_4} + \mathrm{H}, \]

\[ \mathrm{H} + \mathrm{CH_2CO} \to \mathrm{CH_3} + \mathrm{CO}, \]

\[ \mathrm{C_2H_4} + \mathrm{H} \to \mathrm{C_2H_3} + \mathrm{H_2}, \quad \text{etc.} \]

The main product of this photochemical reaction, i.e. of the photodecomposition of ketene under the action of radiation from a high-intensity flash lamp, is acetylene. The formation of carbon occurs either through thermal decomposition of acetylene into carbon and hydrogen, with subsequent polymerization of the carbon,

or by the following reaction:

\[ \mathrm{C_2H_2 + H \to C_2H + H_2,} \]

\[ \mathrm{C_2H \to C_2 + H.} \]

Apparently, carbon is formed in the hydrocarbon flame by the same route.

In an analogous manner, the photochemical decomposition of acetaldehyde, diacetyl, and acetone \(^{22}\), as well as methyl iodide \(^{23}\), was carried out. Acetaldehyde decomposes under a single flash of a pulsed lamp with the formation of carbon monoxide, methane, ethane, and hydrogen. It was found that after the initial photochemical reaction, reactions between the radicals formed predominate (owing to their high concentration), and not reactions of the radicals with molecules of the substance that have not undergone decomposition under the action of light. In particular, the formation of ethane is the principal reaction of the methyl radical. However, in the case of the photochemical decomposition of diacetyl and acetone, both reactions between radicals and reactions of radicals with molecules take place (owing to the low concentration of the former).

Polymerization of acrylates

The production of a large quantity of radicals by irradiation with a pulsed lamp was used for the purpose of polymerizing a number of substances and, in particular, acrylates \(^{24}\). The radicals formed in the flash of the pulsed lamp initiate chain reactions; the product of this reaction has one and the same degree of polymerization, i.e., is more homogeneous in comparison with the polymerization product obtained under continuous illumination. In the latter case, radicals of both the simple monomer and highly polymerized radicals are obtained, as a result of which an inhomogeneous polymerization product is formed. It was found that the rate of polymerization when pulsed light sources are used is directly proportional to the concentration of the free radicals formed and, consequently, proportional to the intensity of the pulsed lamps. In this case it is easy to measure the half-life of free radicals that are no longer in the gaseous state, but in solutions. The half-period of decay of these radicals lies within the limits from 1 to \(10^{-2}\) sec. Thus, for vinyl acetate the half-life of the radical is equal to 0.08 sec.

The examples cited above illustrate sufficiently well the possibilities offered by the new method for studying many photochemical reactions. With its aid, absorption spectra of radicals and atoms were obtained: HS, DS, J, NH, HCO \(^{15,25,26,27}\). However, as yet this method is not so perfected that

register all absorption spectra and, in particular, the absorption spectra of oxygen, chlorine, hydrogen, carbon atoms, etc. Further development of this method with the use of powerful light pulses will probably make it possible to reveal the absorption spectra of the above-mentioned atoms, as well as of polyatomic radicals.

While giving due credit to the advantages of the pulse method for detecting free radicals in comparison with other methods, one should also note its shortcomings: a) it is necessary to consume large quantities of inert gas in order to reduce the thermal effect to a minimum; b) the system does not tend toward a stationary state; c) the course of reactions at high light intensity may differ substantially from that of reactions carried out at low light intensities, owing to the presence in the former case of a high concentration of radicals. In some cases it is possible to detect the presence and even determine small stationary concentrations of radicals from absorption spectra under conditions of continuous illumination. Thus, absorption spectra have been obtained for the radical NH$_2$ with an optical path length of 20 m (with twentyfold reflection) and for the radical CN with an optical path length of 4 m (with fourfold reflection$^{28}$).

II. STUDY OF EXPLOSIVE REACTIONS

Studies of the emission spectra of flames, explosions, and electrical gas discharges have shown the presence in these reactions of about one hundred types of free radicals. At the same time, only those radicals have been detected which are in an excited state. Complex radicals that do not emit light have not been studied. However, from emission spectra it is difficult to trace the rapid changes in the concentrations of free radicals and atoms at the ignition boundary and, consequently, to analyze the entire course of the ignition process. The same may be said with respect to determining the true temperature in these reactions, for temperature, as is known, equalizes with time.

By applying the pulse technique, one can create for these reactions a high concentration of free radicals and atoms and then obtain not only the absorption spectra of the intermediate reaction products, but also measure their concentration during the entire reaction process$^{12,13,18,29}$. The latter is possible only in the case when the reaction begins simultaneously (ignition) throughout the entire reaction vessel. Simultaneous initiation of the reaction (ignition) can occur when the reaction mixture contains a component which, under the action of the powerful flash of a pulsed lamp, decomposes into free

radicals or atoms that are carriers of chain reactions. If the reaction components do not absorb the radiation of the flash lamp, then a photosensitizer is introduced into the reaction mixture, ensuring the initiation of the chain reaction.

Using the flash method, explosive reactions were carried out which may be subdivided into the following types.

Type I—single-component explosive reaction. The radiation of the flash lamp is absorbed by the component itself:

methyl nitrate,
ethyl nitrate.

Type II—two-component system in which the first component absorbs the radiation of the flash lamp:

chlorine—hydrogen,
chlorine—methane,
bromine—hydrogen,
ketene—oxygen,
acetone—oxygen,
carbon disulfide—oxygen,
diacetyl—oxygen.

Type III—two-component system with the addition of a small amount of a sensitizer absorbing the radiation of the flash lamp.

The above-mentioned explosive reactions were carried out at comparatively low pressures of the reaction mixture (from 10 to 20 mm Hg), both with the addition of a large amount of inert gas and in its absence. It was found that the explosive process lasts from \(10^{-3}\) to \(10^{-4}\) sec with a short induction period.

Only explosive reactions of the third type were investigated in detail, namely the reactions of hydrogen and acetylene with oxygen.

Hydrogen–oxygen reaction

As a sensitizer in the hydrogen–oxygen reaction, nitrogen dioxide is used, which, during the flash of the flash lamp, decomposes by half, and the flash \(0.8 \cdot 10^{-3}\) sec—almost completely. The photodecomposition of nitrogen dioxide proceeds according to the following scheme:

\[ \mathrm{NO_2} + h\nu \to \mathrm{NO} + \mathrm{O}, \]

but a side reaction is also possible:

\[ \mathrm{O} + \mathrm{NO}_2 \to \mathrm{NO} + \mathrm{O}_2. \]

As a result of the oxygen–hydrogen reaction sensitized by nitrogen dioxide, absorption spectra of the radicals OH, O₂, NO, NO₂, and HNO₂ are observed. The spectra of one of these radicals, namely the OH radical, recorded at various intervals of time after the flash of the pulsed lamp, are shown in Fig. 7.

Fig. 7. Intensities of the bands of the OH radical 1.0 and 2.0 during the reaction. Hydrogen pressure 10 mm, oxygen 5 mm, and nitrogen dioxide 0.75 mm.

In the course of the sensitized reaction of oxygen with hydrogen, the intensity of the bands in the absorption spectrum changes. They are very intense after the flash, but then rapidly disappear.

If one takes one of the intensities of the bands of the OH radical (Fig. 7) as a standard and compares with it all the other intensities of this same radical, then one can easily construct a graph of the relative concentration of OH radicals as a function of time. All subsequent graphs of the relative concentrations of radicals were obtained in an analogous manner. The concentration of OH radicals in the oxygen–hydrogen explosive reaction also depends on the pressure of hydrogen, nitrogen, and oxygen at a constant pressure of nitrogen dioxide.

Fig. 8 shows the dependence of the OH concentration on time at various hydrogen pressures. The character of the curves indicates

that the rate of decay of the OH radical increases approximately in proportion to the hydrogen pressure. Since the appearance of OH radicals occurs immediately after the photochemical decomposition of nitrogen dioxide:

\[ \mathrm{NO_2}+h\nu \to \mathrm{NO}+\mathrm{O}, \qquad \mathrm{O}+\mathrm{H_2}\to \mathrm{OH}+\mathrm{H}, \]

![Figure 8 graph]

Fig. 8. Relative concentration of OH radicals as a function of time at various hydrogen pressures. NO\(_2\) pressure 2 mm.

the rapid decrease in the amount of OH radicals in excess hydrogen can evidently be represented as follows:

\[ \mathrm{OH}+\mathrm{H_2}\to \mathrm{H_2O}+\mathrm{H}, \]

and in the absence of hydrogen by the reaction

\[ \mathrm{OH}+\mathrm{OH}\to \mathrm{H_2O}+\mathrm{H}, \]

or

\[ \mathrm{OH}+\mathrm{OH}\to \mathrm{H_2}+\mathrm{O_2}. \]

![Figure 9 graph]

Fig. 9. Dependence of the relative concentration of OH radicals on time at various nitrogen pressures. NO\(_2\) pressure 2 mm.

The dependence of the OH concentration on the nitrogen pressure is shown in Fig. 9. Since nitrogen strongly lowers the temperature of the reacting mixture, the formation of OH radicals at high nitrogen pressures proceeds much more slowly. Thus, at a nitrogen pressure equal to 22 mm Hg, the maximum concentration of OH is obtained after \(1.2\cdot10^{-3}\) sec,

whereas without nitrogen this initial reaction lasts less than \(0.1 \cdot 10^{-3}\) sec.

This reaction corresponds to the explosive reaction of oxygen with hydrogen at low temperature. An excess of oxygen increases the lifetime of the OH radicals, and also slightly increases their concentration, i.e., oxygen affects this reaction as an inert gas. The pressure of nitrogen dioxide in all the experiments described above was equal to \(2\) mm Hg. Lowering its pressure, for example, to \(0.6\) mm Hg, leads to an increase in the induction period of the reaction from \(10^{-4}\) to \(2 \cdot 10^{-3}\) sec, as is seen from Fig. 10.

Fig. 10. Dependence of the relative concentration of OH radicals on time at various pressures of nitrogen dioxide. The pressure \((2H_2 + O_2)\) is equal to \(11.25\) mm.

The formation of the radical \(\mathrm{HO_2}\) in these experiments was not detected spectrally, probably because the \(\mathrm{HO_2}\) radical cannot exist for long in the presence of hydrogen, with which it reacts according to the scheme:

\[ \begin{aligned} \mathrm{HO_2} + \mathrm{H_2} &\to \mathrm{H_2O_2} + \mathrm{H},\\ \mathrm{H_2O_2} &\to 2\mathrm{OH} \end{aligned} \]

and, moreover, apparently has no characteristic spectrum in this region.

Acetylene–oxygen reaction

Nitrogen dioxide is also used as a sensitizer in the acetylene–oxygen reaction. By using radiation from pulsed lamps, it was possible to obtain absorption spectra of a whole series of radicals: \(\mathrm{OH}\), \(\mathrm{CH}\), \(\mathrm{C_2}\), \(\mathrm{C_3}\), \(\mathrm{CN}\), and \(\mathrm{NH}\), and even of such radicals whose bands had never previously been observed, for example, the \(3143\) Å band for \(\mathrm{CH}\) and \(4051\) Å for \(\mathrm{C_3}\). These absorption spectra were compared with the emission spectra obtained during the ordinary combustion of a hydrocarbon and showed good agreement.

However, absorption spectra of the products of slow oxidation, such as formaldehyde, could not be obtained.

The initial explosive acetylene–oxygen reaction proceeds as follows:

\[ \mathrm{C}_2\mathrm{H}_2 + \mathrm{O}_2 \to 2\mathrm{CO} + \mathrm{H}_2. \]

The relative amount of free radicals formed depends strongly on the composition of the reacting mixture. In a mixture rich in oxygen, the OH radical is observed together with the CN radical. Increasing the acetylene pressure leads to a decrease in the concentration of the OH radical, and in mixtures rich in acetylene OH radicals are almost absent. As the quantity of OH radicals decreases as a result of secondary reactions, other free radicals appear—\(\mathrm{C}_2\), CH, NH—and an increase in the CN concentration is observed. Thus, for example, the CH radical is formed in the interaction of the OH radical with acetylene:

\[ \mathrm{OH} + \mathrm{C}_2\mathrm{H}_2 \to \mathrm{CH} + \mathrm{H}_2, \]

and the \(\mathrm{C}_2\) radical is formed in the cracking of acetylene.

Fig. 11. Relative concentration of OH radicals as a function of time at various acetylene pressures. Oxygen pressure \(10\) mm and \(\mathrm{NO}_2\), \(1.5\) mm.

All of the above is well illustrated graphically in Figs. 11 and 12. Fig. 11 depicts the dependence of the OH concentration on time at various acetylene pressures (at a pressure of \(\mathrm{C}_2\mathrm{H}_2\), \(p = 14\) mm Hg, the OH concentration is quite small).

Fig. 12 shows how the concentration of the free radicals CN, \(\mathrm{C}_2\), and CH changes with time. In addition, the figure shows that this explosive mixture clearly has an induction period of duration \(0.5 \cdot 10^{-3}\) sec, and this indicates an accelerated development of the explosive reaction and a slow decrease in the concentration of radicals.

In all these experiments the pressure of nitric oxide remained constant, equal to 1.5 mm Hg.

A more complete picture of the acetylene–oxygen reaction is given by Fig. 13, from which it is seen how a change in the ratio of oxygen and acetylene in the reaction affects the concentration of the free radicals formed; moreover, the NH radical appears in mixtures close to equimolecular ones. It is necessary

Fig. 12. Relative concentration of various radicals as a function of time at 10 mm \(O_2\), 13 mm \(C_2H_2\), and 1.5 mm \(NO_2\).

to note that from these curves it is impossible to determine the mutual ratio of the concentrations of the various radicals, since the absorption coefficients of the radicals sometimes remain unknown. The principal final products of this reaction are: with an excess of oxygen—water, and with an excess of acetylene—solid carbon.

Fig. 13. Maximum concentrations of radicals as a function of the pressure of acetylene. Oxygen pressure 10 mm and \(NO_2\) 1.5 mm.

Summarizing what has been said, the kinetics of the explosive reaction may be represented in the form of the following stages:

1. The induction period of the reaction lasts \(10^{-4}\) sec; at its end OH radicals begin to appear.

2. The duration of the second stage is also \(10^{-4}\) sec. During this period, as the result of a whole series of reactions, diatomic radicals OH, CH, and \(\mathrm{C}_2\) are formed; in the presence of nitrogen oxide, CN and NH radicals additionally arise.

3. Reaction between hydrogen and the oxygen residues.

4. Interaction of atoms and radicals, present in high concentration, with one another, with formation of the final reaction products. The most important reactions are the interactions of the OH radical with hydrogen and carbon monoxide.

III. INVESTIGATION OF CHLOROPHYLL BLEACHING \({}^{30}\)

When alcoholic solutions of chlorophyll freed from oxygen*) are illuminated by high-intensity pulsed lamps with a total energy of 390 joules, rapidly reversible changes in the color of the solution are observed. The initial color of the bleached chlorophyll solution returns very quickly and cannot be determined by the methods previously used in studying this phenomenon. The bleaching of the solution and the restoration of its initial color were studied from the change in the intensity of the light of an incandescent lamp passing through a chlorophyll solution that had previously been illuminated by a high-intensity flash. The change in light intensity was recorded by a photomultiplier. After amplification of the photocurrent by a direct-current amplifier, the signal was fed to the vertical plates of an oscillograph; moreover, the light that had passed through the chlorophyll solution was measured not integrally, but by isolating definite regions of the spectrum with the aid of interference filters. Therefore the vertical deflection on the oscillograph showed the change in light intensity that occurs because of the change in the absorptive capacity of the solution in a definite region of the spectrum.

It was found that in the spectral region 4650–4800 Å partial bleaching of the chlorophyll solution is observed; after \(5 \cdot 10^{-4}\) sec it restores half of its initial color. In the wavelength region 5245 Å, after the lamp flash an increase in absorption is observed, and the photoproducts have a greater mean lifetime than the photoproducts detected in the spectral region of 4650–4800 Å. For the wavelength

*) Oxygen prevents the formation of chlorophyll photoproducts in alcoholic solution during the flash of the pulsed lamp.

6310 Å no changes in the absorption of the solution were detected.

On the basis of this investigation it is suggested that, during a flash of high intensity, radicals and chlorophyll molecules are formed in the solution in a metastable, namely, in the lower triplet state.

This supposition is plausible. Academician A. N. Terenin was the first to put forward and substantiate the hypothesis that the metastable electronic state is a biradical triplet state, which determines the course of many photochemical reactions. The absorption of molecules in the triplet state had until recently been observed only in frozen solutions at low temperature. However, if a high intensity of light is used, it is possible to observe the triplet state in solution as well.

Upon illumination by an intense pulsed flash, absorption spectra have been found in solutions of aromatic and polycyclic compounds which correspond to the lowest triplet state itself[^31]. Thus, for example, after illumination of a solution of anthracene in hexane by an intense pulsed flash, along with a decrease in the absorption band of the singlet state, the appearance of new absorption bands at 4203 and 3980 Å is observed. They disappear after approximately 100 μsec, and the absorption of the singlet state reaches its initial value.

CITED LITERATURE

1. U. Waters, Chemistry of Free Radicals, IL, Moscow, 1949.
2. N. A. Prilezhaeva, A. N. Terenin, Journal of Physical Chemistry, No. 8, 111 (1936).
3. G. G. Neĭmin, DAN 16, 453 (1937).
4. V. Gachkovskii, A. N. Terenin, Izv. AN SSSR, OMEN, 805 (1936); Acta phys. chim. URSS 7, 521 (1937).
5. G. G. Neĭmin and B. Popov, ZhETF 5, 87 (1935); B. Popov and G. G. Neĭmin, ZhETF 5, 440 (1935); M. Pavlyuchenko, Journal of Physical Chemistry 14, 605, 877 (1940).
6. V. N. Kondrat’ev, Free Hydroxyl, GONTI, Moscow, 1939.
7. A. G. Gaydon, Spectroscopy and the Theory of Combustion, IL, Moscow, 1950.
8. A. N. Terenin, Proceedings of GOI, issue 40 (1928); Photochemistry of Salt Vapors, GTTI, 1934; Physica 10, 209 (1930); A. N. Terenin and N. A. Prilezhaeva, ZhETF 5, 599 (1935).
9. A. N. Terenin, G. G. Neĭmin, Acta Physicochim. URSS 5, 465 (1935); G. G. Neĭmin, A. N. Terenin, Izv. AN SSSR, OMEN, 529 (1936).
10. R. G. W. Norrish, G. Porter, Nature 164, 658 (1948); G. Porter, Trans. Farad. Soc. 9, 60 (1950).
11. G. Porter, Proc. Roy. Soc. 200, 284 (1950).
12. R. G. W. Norrish, G. Porter, Proc. Roy. Soc. 210, 439 (1952).
13. R. G. W. Norrish, G. Porter, B. A. Thrush, Proc. Roy. Soc. 216, 165 (1953).

1. G. Herzberg, D. A. Ramsay, J. Chem. Phys. 20, 347 (1952).
2. D. A. Ramsay, J. Chem. Phys. 20, 1920 (1952).
3. M. J. Christie, G. Porter, Proc. Roy. Soc. 212, 398 (1952).
4. M. J. Christie, G. Porter, Proc. Roy. Soc. 212, 390 (1952).
5. R. G. W. Norrish, Zeits. Elektrochem. 56, 707 (1952).
6. G. Porter, F. J. Wright, Zeits. Elektrochem. 56, 782 (1952).
7. Martin, Proc. Roy. Soc. 89, 127 (1913).
8. K. Knox, R. G. W. Norrish, G. Porter, J. Chem. Soc. 1477 (1952).
9. M. A. Knan, R. G. W. Norrish, G. Porter, Proc. Roy. Soc. 219, 312 (1953).
10. N. Davidson, T. Carrington, J. Am. Chem. Soc. 74, 6277 (1952).
11. T. G. Majury, H. W. Melville, Proc. Roy. Soc. 205, 323, 496 (1951); G. Oster, H. Mark, J. Opt. Soc. Am. 43, 283 (1953).
12. M. J. Christie, R. G. W. Norrish, G. Porter, Proc. Roy. Soc. 216, 152 (1953).
13. D. A. Ramsay, J. Phys. Chem. 57, 415 (1953).
14. D. A. Ramsay, J. Chem. Phys. 21, 960 (1953).
15. D. A. Ramsay, J. Chem. Phys. 21, 165 (1953).
16. R. G. W. Norrish, G. Porter, B. A. Thrush, Nature 169, 582 (1952).
17. R. T. Livingston, V. A. Ryan, J. Am. Chem. Soc. 75, 2176 (1953).
18. G. Porter, M. W. Windsor, J. Chem. Phys. 21, 2088 (1953).