NEW VIEWS ON AUTO-OXIDATIONS* BY NICHOLAS A. MILAS’
Our constantly increasing knowledge of auto-oxidation reactions since the inception of the Engler-Bach2 theory, has brought forth new but conflicting and at times irreconcilable interpretations of the mechanism of such reactions. Although the Engler-Bach theory gave a reasonable explanation of auto-oxidations, Wieland3 has challenged its validity by presenting new experimental evidence t o prove that oxidations with atmospheric oxygen are not preceded by the addition of molecular oxygen to “auto-oxidants,” as the Engler-Bach theory originally assumed, but by the removal of Eydrogen atoms which are subsequently absorbed by the oxygen molecules to form hydrogen peroxide as the primary product. In other words, Wieland’s theory assumes an initial “dehydrogenation” of the autooxidant followed by an activation of the hydrogen atoms, whereas the Engler-Bach theory assumes an initial addition of oxygen molecules to the autooxidant followed by the formation of “active” oxygen. To explain the numerous autooxidations in which the autooxidant is deficient in hydrogen, Wieland made the assumption that water should add on to the auto-oxidant before it can be dehydrogenated. That Wieland’s contention does not hold with “true” auto-oxidations, is borne out by much chemical and biological evidence. It will be beyond the scope of this paper to attempt to give a complete list of such oxidations; only a few representative examples will suffice to show the inadequacy of Wieland’s view. Lothar Meyef was the first to observe that rigidly-dried mixtures of carbon monoxide and oxygen could be made to combine non-explosively, while the more recent and more reliable experiments of Bone and his students5 proved beyond any reasonable doubt that “the presence o j steam is not essential to the ignition and explosion of carbonic oxide and oxygen mixtures, &s has been supposed hitherto,6 but that the two gases can and do combine
* Contribution from the Research Laboratory of Organic Chemistry, Massachusetts Institute of Technology, No. 43. ‘Research Associate, M.I.T. The greater part of this paper was written while the writer served as a National Research Fellow a t Princeton University (1926-1928). * Engler and Wild: Ber., 30, 166g (1897); Engler and Wekberg: “Kritische Studien uber die Vorgange der Autoxydation,” (1904). Bach: Compt. rend., 124, 951 (1897); “Fortschritte d. naturw. Forsch.” (Edited by Abderhalden), 1, 85 (1910). JWieland: Ber., 45, 484, 685,.679, ?603 (1912);46, 3327 (1913);47, 2085 (I914), and several other articles up to and mcludmg 1928. Meyer: Ber., 19, 1089 (1886). 5Bone, Newitt and Townend: J. Chem. SOC., 123, 2008 (1923); Bone, Fraser and Newitt: Roc. Roy. SOC.,llOA, 615 (1926); Bone and Townend: “Flame and Combustion in Gmes,” 340 (1927). The authors refer, in this case, to the early experiments of Dixon: British Association Reports, p. 503 (1880);Phil. Trans., 1 7 5 , 6 1 7 (1884); J. Chem. SOC.,4 9 , 3 8 4 (1886). These experiments have been recently cited by Adickes: 2. angew. Chem., 40, 1131 (1927), as conclusive evidence in favor of Wieland’s dehydrogenation theory.
NEW VIEWS ON AUTO-OXIDATIONS
I205
directly without its intervention.” Moreover, Lind7 reports that under ionizing influence carbon monoxide combines with oxygen even at the temperature of liquid air and concludes that “reaction a t this low temperature is a proof that moisture is not essential to the propagation of the ionic reactions.” Similar conclusions have been reached by Dixons in the case of the combustion of dry mixtures of cyanogen and oxygen, and by Bakerg and by Dixon and RussellIo in the case of the combustion of dry mixtures of carbon disulfide and oxygen. Furthermore, MellorII states that extreme drying does retard the oxidation of phosphorus, but does not completely inhibit the process. One finds equally striking chemical evidence in favor of the Engler-Bach theory in the successful attempts made recently to isolate “true” organic peroxides from auto-oxidation reactions. Clover1* succeeded in isolating ether peroxide which, according to him, is responsible for the explosions occurring during distillations of ether. Jorissen and van der Beek13 have recently prepared benzoperacid in relatively good yields by the auto-oxidation of benzaldehyde. Very recently Stephens14claims to have isolated a “true” peroxide of cyclohexene. while MardlesIs succeeded in demonstrating the primary formation of extremely active and short-lived organic peroxides in the slow combustion of ethane. The experiments of Mardles are of special significance since they are at variance with the “hydroxylation theory”16 of hydrocarbon combustion. Moreover, the very recent experiments of Mila# on the induced oxidations effected during the auto-oxidation of anethol, styrene, etc., demonstrate clearly the primary formation of unstable but very reactive peroxides. Another source of evidence is to be found in the field of inhibition. The recent observations of Dhar,I8 Moureu and Dufraisse,Ig and others have shown that small quantities of certain inorganic as well as organic substances are capable of retarding or even preventing the oxidation of auto-oxidants, such as acrolein, benzaldehyde and sodium sulfite with molecular oxygen. Furthermore, Dixonm has observed that small quantities of ethylene and
’
Lind: “The Chemical Effects of Alpha Particles and Electrons” (Second Edition) 167 (1928). Dixon: J. Chem. Soc., 49, 390 (1886). OBaker: Phil. Trans., 179, 582 (1888). Io Dixon and Ruasell: J. Chem. Soc., 75,600 (1899). Mellor: “A Comprehensive Treatise on Inorganic and Theoretical Chemistry,” 8, 775 (1928). l2 Clover: J. Am. Chem. Soc., 44, I 107 (1922). I 3 J o r k e n and van der Beek: Rec. Trav. chim., 45, 245 (1926);van der Beek: 47, 286 (1928). Stephens: J. Am. Chem. Soc., 50, 568 (1928). I5 Mardles: J. Chem. Soc., 1928, 872. I6 Bone and Townend: 1.c. page 373. I’ Milas: Proc. Nat. Acad. h i . , 14, 844 (1928). Dhar: Proc. K. Akad. Wetensch., Amsterdam, 29, 1023 (1921). Moureu and Dufraisse: Institut Internationale de Chimie Solvay, Bruxelles, April (19251,page 524; Chem. Reviews, 3, 113 (1927). *O Dixon: Rec. Trav. chim., 4‘2, 3 0 j (1925).
1206
SICHOLAS A. MILAS
acetylene have a powerful inhibitory action on the phosphorescent combustion of carbon disulfide. In the field of biology, Warburg?’ made the significant obserration that small quantities of narcotic?? or poisonous substances not only impede the respiration of living cells but inhibit it in a “reversible manner.” Since the early work of Karburg important observations have been made in t,his field by M e y e r h ~ f , *Fleisch,?l ~ Szent-Gyorgyi,?S and On the basis of the foregoing experimental evidence, one may reasonably conclude that, with very few exceptions, one important criterion of auto-oxidations whether in gaseous, liquid or solid state,*; in ciz’o or in z’iti-o, is their susceptibility to inhibitory action. Wieland’s theory is incapable of explaining these experimental facts since dehydrogenations are very little or not at all susceptible to inhibitory action.28 Still another source of evidence in favor of the “peroxide theory” is to he found in the fields of promoter action and induced oxidations. T h ~ n b e r g ? ~ was the first to show that small quantities of iron salts produced startling effects on the auto-oxidation of lecithin. I n this connection W a r b ~ r gMey,~~ erhof31and others have further demonstrated that iron salts are capable of accelerating auto-oxidations even in the living cell, while Spoehr3*and Spoehr and Smith33have shown that sodium ferro pyrophosphate is an excellent oxygen carrier in inducing the oxidation of glucose and related sugars with atmospheric oxygen. This catalyst was found to induce not only the oxidation of reducing sugars but also that of sucrose and polyhydric alcohols which do not reduce Fehling’s solution. The authors are not quite convinced as to the exact mechanism of their reactions but point out three possible ways whereby oxidations may occur, either by means of surface catalysis (adsorption), or by electronic oxidation and reduction of the catalyst! or by intermediate peroxide formation with the catalyst. An important class of oxygen carriers whioh seem to play a very significant rBle in respiration phenomena of the living cell is that which includes compounds of the sulphydryl group which was first suggested by the work of 21
\Tarburg: 2. physiol. Chem., 69,4j2 (1910); Erg. Physiol., 158, 190 (1914);Biochem’
z., 136,266 (1923).
22 Winterstein: “Die Xarkose,” 214, (1926)states that inhibition of auto-oxidations in the living cell should not be considered a specific action of narcotics and therefore is not the cause of narcosis. 23 Meyerhof: “Chemical Dynamics of Life Phenomena,” Chapter I (1924). 24 Fleisch: Biochem. J., 18, 294 (1924). Szent-Gyorgyi: Biochem. Z., 150, 19j (1924);157, jo,298 (192j). 26JVind: Biochem. Z., 159,58 (1925). 27Feldmann: Giorn. chim. ind. applicata, 9.455 (1927). 28For a crucial test of Kieland’s theory, see Tanaka: Biochem. Z., 157, 42j (192j’; J. Oriental bled., 4, 4 (1925). Thunberg: Skand. ;Irch. Physiol., 24,90,94 (1910). 3oWarhurg: Biochem. Z., 142,jI8 (1923);145,461 (1924);152,479 (1924). 31 Meyerhof: 1.c. pages 17-22. 32 Spoehr: J. Am. Chem. Soc.. 46, 1474 (1924). 33 Spoehr and Smith: J. Am. Chem. Foe., 48,236 (1926). *Q
XEW VIEWS O S AUTO-OXIDATIOSS
1207
Heffter.34 Glutathione3: and thioglycollic acid36are two compounds of this group that have been extensively studied both an tztro and in L ’ Z L O . Various mechanisms have been proposed to account for the ability of these compounds to transport oxygen to muscle tissue. Some biologists believe with Kieland that the sulphydryl group does not actually transport oxygen to the muscle tissue but instead removes hydrogen atoms from it. The assumption is made, therefore, that the disulfide which is formed by the spontaneous osidation of sulphydryl acts as a hydrogen acceptor. That this mechanism is not plausible is clearly pointed out by Neyerhof who states that, “According to this theory (dehydrogenation) it remains, however, incomprehensible, how the sulphydryl compound can transport the oxygen in a larger amount and above all with greater speed than the disulfide.” He then postulates a mechanism whereby oxygen may be transported to the muscle tissue by the sulphydryl :
0-0 HSH
+ HSR + Oi+RSH
j
HSR
0-0
I 1
RSH
HSR + lI---t2RSH
+ A102
where izI represents the muscle tissue. Obviously this interpretation gives a more plausible explanation of the facts, but chemically it is highly improbable. More recently Kendal13’ suggested an interpretation of this process in which he made the assumption that the ability of glutathione and other analogous substances to transport oxygen to muscle tissue depends on the presence of “thermolabile unstable addition products containing oxygen and sulfur.” ;Inother interpretation of this process based on the recent advances of electronic structures is proposed later in this paper. Auto-oxidations are not only susceptible to inhibitory and accelerating action of certain types of reagents, but are also capable of inducing the oxidation of other substances which ordinarily are not osidised with molecular osygen. As far back as 1890 Pedle+ noticed the oxidation of atmospheric nitrogen during the combustion of carbon disulfide with air, while very recently ?rIardles15 reports the induced oxidation of benzene, aniline, etc., during the slow combustion of ethyl ether or ethane. Furthermore, Feldmann?’ has recently shown that sodium amalgam induces the osidation, with molecular oxygen, of mercury to mercuric oxide a t rooo, also that of ethyl alcohol to acetic acid under pressure. Xeither the mercury nor the alcohol alone are readily oxidized with oxygen under the conditions stated. The Heffter: 3Ied. nat. Arch., 1, 81 (1908). Hopkins: Johns Hopkins Hospital Bull., 3 2 , 321 36 3Ieyerhof: 1.c. pages 30-35. 3i Kendall: Science, fz), 62, 384 (19281. Pedler: J. Chem. SOC..57, 6 2 j (1890). :-I
(1921);
Biochem. J., 15, 286 (1921).
NICHOLAS A. MILAS
I 208
present author’’ has further shown that during the auto-oxidation of anethol, styrene, etc., several substances like anthracene. phenanthrene. etc., are inducedly oxidized. All these substances have been shown to be quite inert to oxygen in the absence of the auto-oxidant. Therefore, one may safely conclude that another very important and quite general characteristic of auto-oxidations is their ability to induce the oxidation of other substances. I n accordance with Wieland’s view these reactions remain utterly incomprehensible. The critics of the Engler-Bach theory have advanced the argument that organic peroxides alone, active though they may be, are not sufficiently strong oxidizing agents to oxidize either the muscle tissue or other comparatively inert organic substances. In order to test the justification of this criticism, the author performed the following experiment: Two g. of anthracene (Eastman Kodak, C.P. quality, free from anthraquinone) was suspended in 40 C.C. of glacial acetic acid contained in a 1 2 0 C.C. test tube and to the mixture was added 5 g. of b e n z ~ p e r a c i d . ~The ~ test tube with its contents was placed in an oil-bath kept at IIOT for one hour. The experiment yielded, besides other oxidation products of anthracene. 0.25 g. of anthraquinone. m.p. 281283’C. This experiment seems to prove the inadequacy of the foregoing criticism. However, the peroxide theory, alone, cannot, in its present form, adequately explain the induced oxidations. Neither is it competent of explaining satisfactorily, in the opinion of the author, the retardation or acceleration effects produced on auto-oxidations by small quantities of certain types of chemical substances.
The New Interpretation Granting that in all auto-oxidations oxygen adds in the molecular form, is there any new evidence to show how this addition might take place? This rather fundamental question may be answered satisfactorily only after a thorough and critical examination of the most recent views concerning the possible electronic configuration of oxygen on the one hand and of autooxidants on the other. There seems to be a confusion, a t present, as to the exact electronic configuration of molecular oxygen. On the basis of the paramagnetism of oxygen, one may conclude that the oxygen molecule which has a moment corresponding to two Bohr magnetons may be analogous to an atom with two electrons rotating in the same direction.40 Rut all gaseous substances exhibiting paramagnetism consist of molecules possessing an “odd” number of electrons and oxygen therefore constitutes a distinct anomaly. To explain this anomaly Lewis4‘ made the assumption that the predominant form of molecules in free oxygen is that in which each oxygen atom possesses an “odd”
..
..
electron as, : 0 : 0 :. Accordingly Taylor and Lewis4*thought that if such a Houben-Weyl: “Die Methoden der organischen Chemie,” 2, I 13 (1922). Stoner: Phil. Mag., (7), 3, 336 (1927). Lewis: Chern. Rev., 1, 243 (1925). 42 Taylor and Leais: Proc. Sat. .lead. Sci., 11, 456 (1925).
40
4L
NEW VIEWS ON AUTO-OXIDATIONS
I 209
molecule reacted with atomic chlorine, one of the “odd” electrons should either be transferred to the chlorine atom or be shared with it SO that an inert configuration is formed around the chlorine atom. The remaining “odd” electron on the other oxygen atom should then impart paramagnetic properties to the chlorine dioxide molecule. This prediction was confirmed when the authors found that C102 in CC14 was, indeed, paramagnetic. For a time, this evidence was quite convincing. Nevertheless, the anomaly still remains a mystery. Very recently Grimm” proposed an electronic structure for C102 analogous to that of Nos,
.. .. C1 : 0 : . . . . . .
10
..
:O
..
. . .
N
0:
..
Although Grimm’s structure may account for the paramagnetic properties of CIOz, it fails to explain the radical difference in chemical properties between the two gases. Strange to say, both Lewis and Grimm failed to consider other possible structures. T’ierefore, if we make the following assumption that in all auto-ozidations (oxidations in which oxygen reacts in the molecular form) the atoms to which the ozygen molecule initially adds m a y be regarded as making definite cuntributiuns of two electrons to i t , in the case of the chlorine atom, the following electronic structure of chloride dioxide may be deduced,
..
.c1: ..
+
. . . . 0 : 0: . . . .
. . . . . .
.c1 . .: .0. : . 0. :
:
This then, may be termed “dative” peroxideu of chlorine. It seems to explain not only the paramagnetic properties of the gas, but also its instability and explosive nature under the influence of light or heat. In accordance with this reasoning, we are justified to conclude that molecular oxygen reacts by sharing two electrons with other molecules t.0 form metastable or dative peroxides45which are characterized with high instability and energy contcnt. Owing to their high instability, these peroxides may either revert instantaneously to the ordinary peroxides by transferring their excess energy to other molecules, or initiate the oxidation of other molecules.
Molecular Valence Electrons On the basis of the reasoning in the preceding section, it may be inferred that auto-oxidations can occur only when the auto-oxidant possesses unshared or “exposed” electrons comparable to valence electrons of the various Grimm: “Handbuch der Physik,” 24, 513 (1927). The term “dative” has been first suggested by Menzies: Sature, 121, 457 (1928),to denote a type of covalency in which one of the atoms in the “dative” bond contributes both electrons. 45 It is best here to point out that in developing this view of autwxidations, I have derived ‘yonsidersble benefit from the recent views of Sidgwick: J. Chem. SOC.,123, 725 (1923); The Electronic Theory of Valence,” 116 (1927). 45 Shared electrons which can be easily “exposed” by absorption of some form of external energy. 43 44
SICHOLAS A . MILAS
I210
element^.^' I n view of the recent developments in connection with electronic bands, ?\Iulliken4sand Birg@ have shown that the energy levels associated with valence electrons of molecules are analogous in all essential aspects with the valence electrons of atoms. In other words, molecular valence electrons are capable of existence in a series of electronic states. I n molecules like CO, for example, it is assumed that each atom retains its own K electrons, and of the outer ten electrons, eight form some sort of a firmly bound symmetrical group, while the remaining two are more loosely bound and therefore give rise to a set of energy levels which are analogous t o those attributed to valence electrons in atoms. These two electrons which constitute a “lone“ or unshared pair, are termed “molecular valence electrons.”
Kow, if the view expressed in the early part of this section is correct, one would naturally expect CO to combine with one molecule of oxygen to form an unstable and highly reactive dative peroxide which will further react with another molecule of CO to form two molecules of CO?, as
..
. . . .
..
. . . .
o : c : + o : o : - + : o : c : o : o .. . . . . .. . . . . ..
. . . .
..
: o : c : o : o : + : c : o :-+sco2. ..
. . . .
..
The esplosive nature of this reaction, as noted by Bone and his students,j indicates clearly the possibility of an intermediate metastable peroxide formation. This view may be extended further to include unshared electrons found in molecules like S O , SH3, PH,, SO?, PHSOX, XHJHZ, H?S, H I and thow containing unsaturated linkages. Xll these substances are auto-oxidized, although under different conditions, a fact which seems to indicate that t h e molecular valence electrons are a t different “penetration” levels depending upon the “promotion energies,’’ the effective nuclear charge of the parent atonPo and the stability of the “unprornoted” shared electrons in the autooxidant molecule. For example, XH, reacts with 0 2 only at relatively high teniperatures, while PH3 burns spontaneously in 0 2 even at the temperature of solid C‘O,. This difference in reactivity between S H , and 0 2 and PH3 47 Since the writing of the present paper. a very important article has been published I,y 3Iulliken: Phys. Rev., 32. 186 ihugust 1928!, in which the \Triter uses the term “promoted” electrons to denote elertrons 5~hoseT L values (principal quantum numbers) have l)een increased as a result of atomic combinations and the formation of molecules. Nulliken: Phys. Rev., ( a ; 26, 561 (1925); Proc. S a t . dead. Sci., 12, 144, 338 ( 1 9 2 6 ) . 4 9 Birge: Bull. S a t . Res. Council, 11, part 111, 69 (1926). bo Latimer and Rodebush: J. Am. Chem. Soc., 42, 1419 11920).
S E W VIEWS O S AUTO-OXIDATIOSS
12x1
and O2 indicates not only that the two molecules have different’ degrees of “penetratiod’ of molecular valence electrons but also a difference in energy content of the more firmly bound electrons. But is there any chemical evidence to show that molecular oxygen combines initially with substances containing molecular valence electrons? From the auto-oxidation of (C2Hj)3P,Jorissenj’ succeeded in isolating (C2H5)3Pb, the peroxide properties of which have been demonstrated by Engler and Wild.32 Moreover, one might expect that the “blocking” of molecular valence electrons by other atoms would make the molecule more inert to oxygen. This is actually found to be the case with several substances. In Table I, the compounds listed under Column I are readily oxidized with oxygen, while their salts listed in Column I1 are relatively stable in oxygen.
TABLE I substances easily osidized
H3P:j3
Substances stable in oxygen
H3P:H I
(CZH5)3P :”
(C2Hj)3P:C2H5 Br
(CzH5)3Sb :”
(C2HJ3Sb:C2H5 Br
..( C H ..~ ) ~ X S : ”
(CH3)3;1s:CH3 Br
KH?KH2’6
SH2?;H2(H?SO4)2
R%H?H2’’
R S H SHZHZSOI
ArKHOHss
ArSHOH H2SOI
In addition to these substances several aryl amines, hydrazones, etc.. ell known to organic chemists, are isolated only in the form of their salts. In accordance with the view presented in the foregoing pages, one would exppct the substances on the right hand column of Table I to react with oxygen by virtue of their molecular valence electrons. Since t,he degree of reactirity of these substances Iyith oxygen is known only qualitatively a t pre$ent, we cannot draw any definite conclusion other than to state that the molecular valence electrons present in these substances may be insufficiently “energized” to initiate reaction with osj-gen under ordinary conditions. Jorissen: Ber., 29, 170; i1856’1. Engler and JYild: Ber., 30, 1673 (189;). 53 Olscheivsky: Nonatshaft, 1, 3 7 2 :1886,. Christiansen: “Organic Derivatives of Antimony,” p. 2 j (192jl. j 5 Raizis and Gavron: “Organic Arsenical C‘ornpounds,” p. 66 1923 5 b Cuy and Bray: J. h m . Chern. doc., 46. 1786 (1924). 5’ Chattaway: J. Chem. Soc., 91. 1323 (19071, j8 Bamberger: Ber., 27, I j j I :18941; 33, 118 (1506:. 51
5*
NICHOLAS A. MILAS
I212
We are now in a position to explain the ability of sulphydryl groups to transfer “active” oxygen to muscle tissue or to other organic substances.
.. : o : ..
: o : .. R
..
..
..
: S : H + O : 0 : -+ R : S : H
..
..
..
..
R : S :H + M
..
.. -+ R
:S: H+MO*
..
..
: o :
.. : .o. : Similarly, the ability of the disulfide to transfer the oxygen may be explained :
.. : o : ..
: o : ..
..
.. R : S : S : R + M
-+
.. . .
..
R : S : S :R+MOp
.. ..
However, the “penetration” of molecular valence electrons in the disulfide is much greater, due to a considerable liberation of energy during its formation from the sulphydryl, and therefore it would be expected to be a poorer oxygen carrier. The absorption of molecular oxygen by unsaturated groups like -C=C,
H I
> C = C < , -C=O,
H +\
> C = O , +-C-,
+/
I
-C=X-,
> C = S , etc., to form
dative peroxides may be best illustrated by the following two typical esamples :
1213
N E W VIEWS ON AUTO-OXIDATIONS
( I ) Auto-oxidation of Styrene:-
H
H C
H
C
/\ .. . . CH il I f 0 : o : .,. . . HC CH \/ H c .. HC CH ..
/\ CH 11 I
HC
HC
+
HC
CH H
\/ c
HC
:
.. HC
CH ..
.CH2 .
..
: o :
: o :
..
..
0 :
..
: o : ..
H C
H C
/\ CH I/ 1
HC HC
CH
+ Os(Active),
HC
CH
HCH
+ /I
0
\/
\/ C HC
/\ CH 1; I
HC
:
C HC=O
CH2
In the first case the active oxygen will oxidize other substances inert to molecular oxygen. Of course, the aldehydes formed during the second course of the reaction oxidize further to the corresponding acids. ( 2 ) Auto-oxidation of Sulfones :-De16pine59 observed the formation of ketones and sulfur trioxide in addition to small quantities of sulfur dioxide and sulfur. The main reaction may be expressed as follows:
..
“>c: s : + ..
R
58
..
..
0 : 0 :
..
Delepine: Bull., 31, 762
..
(1922).
+
I‘>.:
.. s :
..
R
:
0 :
:
0 :
R\C--S
+A
0-0
I
SICHOLAS A . JIILBS
1214
0--0
+
0-I-0
R>=0+s03 R
Side reactions are also likely to occur, due, perhaps, to the interaction of PO., with the original sulfone molecules. If the mechanism, as shown above, is correct, the double dative peroxide represented in the second reaction, mould possess much more energy than would a single dative peroxide. This is born? out by the resuhs of Delepine who observed that during the spontaneous auto-oxidation of sulfones considerable visible radiation is produced. The spontaneous auto-oxidation of yellow phosphorus may he analogously interpreted. Inhibition and Acceleration of Auto-Oxidations In the interpretation of band spectra, modern spectroscopists have arrived at the conclusion that the total energy of a molecule is made up of the energy due to the transitions of electronic orbits, the oscillatory energy of nuclear vibrations and the rotational energy of the molecule as a whole. Since the molecular valence electrons are niore loosely hound to the molecule, they may be assumed to be more easily affected by environmental disturbances, such as collisions from neighboring molecules or elect’rons,radiation and ternperature effects. Therefore, the first change ~ c h i c hm a y occur in a n y chemical reaction i s the change of electronic energy caitsed by eni’iTonrnenfa1forces. This change causes an inirnediate change in the vibrational and rotational energies of the molecule in order that equilibrium may be established. Arrhenius was the first to point out that molecules should be in an “activated” state before they can take part in any chemical reaction. In the case of auto-oxidants, the “activation” or the “de-activation” of molecules may be governed by the following energy considerations: ( I ) The existence of activated molecules depends upon the “penetration” of molecular valence electrons. Penetration may be defined as the energy level of the molecular valence electrons at, a definite equilibrium state of the molecule in question. ( 2 ) The energy transfer among molecules under ordinary conditions is manipulated entirely through changesof the loosely bound molecular valence electrons. If A : represents a molecule of an auto-oxidant and el the energy of activation, the changes which may occur during the auto-oxidation of the molecule .A: may be represented as follows: A:
+ el + A > :
(1)
NEW VIEWS O S AUTO-OXIDATIOSS
1215
where e2is the energy due to the instantaneous neutralization of the activated molecule .A 2+: and 0 2 . Furthermore, the electrons which are directly responsible for the formation of the dative peroxide shown in ( 2 ) have undergone an increase in their principal quantum numbers due to the absorption of energy e2. Molecules containing these electrons would be characterized with a very low energy of dissociation and extremely high instability. Mere collisions with other molecules therefore will suffice to effect a rearrangement or even decomposition of the dative peroxide with liberation of energy. This energy, according to Christiansen60is utilized to initiate reaction chains.
+
where e3 is the sum of el e2. This energy increases with each subsequent reaction chain and eventually, according to Semenoff .61 leads to explosive reactions.
If an inhibitor collided with the dative peroxide, all the excess energy of the latter would be completely absorbed by the molecular valence electrons of the former, thus the initiation of new reaction chains is prevented and the rate of auto-oxidation is greatly reduced. A critical examination of the various inhibitors of auto-oxidation reveals the fact that all of them possess molecular valence electrons, but probably at different penetrations. An exchange of energy, therefore, between the inhibitor and the dative peroxide resulting in the partial activation of the former, is inevitable to inhibitory action. hloreover, this exchange of energy takes place instantaneously either by electronic impacts or by emission and re-absorption of invisible radiation. The activated molecules of the inhibitor can then be either oxidized by the organic peroxides or by free oxygen,62or combine with the active auto-oxidant molecules forming unstable complexes which may decompose to yield the original inhibitor molecules. Anthraquinone, for example, was found by the author to inhibit the auto-oxidation of anethol yet it was recovered unchanged at the end of the reaction." The mechanism of the promoter action on auto-oxidations, on the other hand, can be explained by assuming that the valence electrons of the promoter transfer their energy to the inactive auto-oxidant either directly or through the formation of an intermediate dative peroxide of the promoter. Christiansen: J. Phys. Chem., 28, 145 (1924). Semenoff: 2. physik, 48, 5 7 1 (1928). a Backstrom: Medd. K. Vetenskapsakad. Sobel-Inst., 6,38 (1927). 63 Xilas: Proc. Kat. Acad Sei, July (1929). 6o
1216
NICJTDLAS A. MILAS
Schematically, the inhibitory and accelerating action produced by small quantities of substances on auto-oxidations and the induced effect of the latter may be explained as follows:
A :
+ el
-+
-A% :
A 2 :
(1)
.. ..
.. ..
+ 0.. : 0.. : + A : 0.. : 0.. : (el + e2)
(2)
(a) A : +BO*
(b) A = O
+ B=O
Reaction (b) is irreversible, while (a) might under favorable conditions, reverse itself. In the case of the promoter action, the following scheme may be adopted:
A:
+ C;
: -+ A 2!+:
+ C : (Promoter action)
.. .. .. .. (2%: + .. 0 : .. 0 : - + C : 0 : 0 : (ei + e*), .. ..
or and
(4)
.. ..
+ e * ) + : A -+ C : + A02 + el + es or + G :* C : + AOz + zel + ez (5) Reaction ( 5 ) shorn promoted + induced oxidation. In the above scheme B C :0 .. : 0.. : (el 7,
represents the inhibitor and C the promoter. As a concluding remark, it may be stated that the new interpretation of auto-oxidation reactions, as presented in the foregoing pages, though it differs from any views so far presented serves, in the opinion of the author, as a compromise among the prevalent, widely different views held by the various investigators in the field. Research Laboratory’oj Organw Chemistry, Massachusetts Institute of Technology, Cambrzdge, Masa. March SO, 1929