ACTOXIDATION RETARDATION BY POLYARYLMETHANES
June 20, 1964
237 1
Reagents.-Research grade cyclopentene (99.89 mole yG), oxide by the procedure of Barnard arid Hargrave.lg In this cyclohexene (99.98 mole 7G), and cumene (100.0 mole yG)were analysis nitromethane interfered slightly and a blank correction obtained from Phillips Petroleum Co. Cycloheptene was preobtained using the same amount and concentration of nitropared by the reduction of cycloheptanone by lithium aluminum methane was applied. Table Y summarizes typical hydrohydride followed by elimination of water catalyzed by naphthaperoxide yields. I t is not believed t h a t the solvents employed lenesulfonic acid. Cyclooctene and cycloocta-1,5-diene were (Table I ) were involved in a cooxidation sequence. In the gifts from Cities Service Research and Development Co., while absence of cyclohexene or cumene the solvents themselves did cycloheptatriene and norbornadiene were gifts from Shell Cheminot undergo oxidation a t a significant rate. Moreover, many of cal Co. Indan (Aldrich Chemical C o . ) and tetralin (Matheson, the solvents would be inert t o peroxy radical attack because of Coleman and Bell) were shaken with sulfuric acid, washed, and the polar effect of electron-~~ithdrau.ing functional groups such dried. Cyclohexa-l,3-diene was prepared by bromination of as nitro, cyano, sulfoxide, carbonyl, and carboxyl substituents. cyclohexene with N-bromosuccinimide followed by quinolineSubstituents such as ethers, sulfides, amines, or amides were catalyzed dehydrobromination.20 The preparation of p-nitronot employed because of the possible complications from cocumene has already been described." oxidation. The hvdrocarbons other than those of research grade mere TABLE V rectified in a Todd column with a m o d spiral and center Is~ ~ U T O X I D A T I O SOF cYCLOHEXESE~ fractions selected having constant boiling points and having no L;C of absorbed impurities detectable by gas chromatography. The purified 02 found a s hydrocarbons were passed through silica gel under a dry nitrogen Hydrocarbon Solvent ROOH atmosphere and stored under nitrogen prior to use. CycloCyclohexene CHzSOz 101 hexenpl hydroperoxide was prepared by oxidizing a large volume (CHJ)?CHOH 104 Cycloheuene of cyclohexene to the extent of a few per cent with AIBK and Tetralin CHI SO^ 93 oxmen. _ - The excess cvclohexene was removed a t reduced DresCycloocta-l,%diene CH3T02 90 sure and the hydroperoxide fractionated in vacuo. Purity by hydroperoxide analysis was 82.57'. Cycloheptene CH3S02 65 Solvents.-Commercial solvents were distilled prior to use Cycloheptatriene CHaS02 47 and where possible passed through a silica gel column prior t o use. Cycloheptatriene CsHbC1 36 t-Butylnaphthalene was prepared by the reaction of t-butyl alcoC~clooctene CsHsCl 62 hol with naphthalene in the presence of boron trifluoride and CycIopenteneh CHzS02 70 phosphoric anhydride. Distillation gave material, b.p. 84-89' Sorbornadiene CHzS02 9 (1 m m . ) , n Z o1.5694, ~ which was passed through silica gel before a 2 0 JI hydrocarbon a t 60" in presence of 0 0504 M AIBK. use. * 1 0 -11hydrocarbon ~
(19)
n. Barnard
~~
120) G S H a m m o n d and J Warkentin (1961)
a n d K Hargrave, A n a ! . Chem. A c l ~ 6, , 476 (1951).
[COSTRIBUTION FROM
THE
DEPARTMENT O F CHEMISTRY,
IOWA STATE
J A m Chem S O C ,83, 2554
UNIVERSITY, AMES,I O W A ]
Retarding Effects of Polyarylmethanes in Autoxidation Reactions1a2 BY DALEG. HEN DRY^^
AND
GLENA.
RECEIVED J A N U A R Y 6, 1964 The retarding effect of polyarylmethanes or polyarylalkenes on the autoxidation of cyclohexene or cumene has been established to be due to the low reactivity of the polyarylmethyl radicals formed by attack of peroxy radicals. The polyarylmethyl radicals can persist in the presence of oxygen and retard the reaction by trapping peroxy radicals to give nonradical products. Abilities of polyarylmethanes to inhibit autoxidation reactions parallel and are a measure of the resonance stabilization of the polyarylmethyl radicals. Inhibition of this type becomes more pronounced as the temperature is raised.
The rate of autoxidation is generally retarded by materials which destroy the chain-propagating free radical intermediates. The effectiveness of phenols and anilines as inhibitors depends on the reactivity of these substances with peroxy radicals and the inability of the resulting phenoxy or anilino radicals to carry on effectively the chain ~ x i d a t i o n . ~ The presence of substances which give rise to peroxy radicals which readily terminate can bring about a retardation of the autoxidation of substances which yield peroxy radicals that terminate a t a slower rate.5 Thus sniall amounts of tetralin drastically retard the oxidation of cumene even though tetralin is more reactive than curnene toward peroxy radicals.jB6 I n this competitiL-e oxidation it can be proved t h a t only peroxy radicals are involved in the termination process.5 11 1)iicrtive E f f e c t s in Aliphatic Suhstitutions XXVII. ( 2 ) 'l'his work was supported by a grant f r o m t h e Petroleum Research Pounilatiori adminiiter-ed by t h e American Chemical Society Grateful arhncruledyment is heieby made t o the donors of this filnd 1 3 1 la) Ethyl Cilrpoi-ation PI-edoctoral Reseal-ch Fellow, 1960-1961 If]-ed P Sloan Foiinrlati triphenylmethane > a-naphthyldiphenylmethane > 1,Ldiphenylethylene > triphenylethylene. Table I 1 also shows that the retarded rate of oxidation is dependent on oxygen pressure and is lower a t a lower oxygen pressure. Temperature Dependence of Retardation-The results of cooxidations of cumene and triphenylmethane performed a t 60 and 90' are summarized in Fig. 2 . The retardation is more pronounced at 90 than a t 60'. This is somewhat surprising since one would expect t h a t triphenylmethane would be attacked more selectively a t the lower temperature. The effect of ternperature on retardation suggested that possible oxygen reacted with the polyarylmethyl radicals rapidly but that the reaction was reversible. T o test for this possifast
+ I-iZO?
-1. RO2.
0 2
+ .-I.---+
ROn.4
bility we searched for a reagent stable in the presence of oxygen which would react readily with XO2, and which would propagate the chain reaction by regenerating R and A .. (12) D. Barnard and K. R. Hargrave. Atla1 C h i m A r t n . 5 . 4 (13) I,. Bateman. Oirrirl. H a (1,ondsn). 8 , 1-17 (19%) (14) G A. R u ~ s e l l .I, A m f h e m S O L ,79, 3 9 1 ilD.57) (15) G. S . H a m m o n d , J . S Sen, and C E. I3oozer. ihid , 77, ( I S ) J . P. Van Hook and A V, T o h u l s k y . i h i r i , 8 0 . 779 (i%j8)
(l!IATI
RETARDATION BY POLYARYL3IETHANES
June 20, 1964
23i3
A~UTOXID.ATION
+ I-, I' + R H (or AH) --+ H T + R . (or .A,) AO,
+ ST -+-
AOLT
The only reagent we know of that can perform the requirements demanded of x - U is hydrogen bromide. l 7 The oxidation of cumene in the liquid phase containing traces of hydrogen bromide proceeds extremely rapidly initially b u t soon inhibits itself presumably due to the TABLE I1
COOXIDATIOV OF CVCLOHEXENE ASD POLYARYLMETHANES~
Added hydrocarbon
Sone Xone Triphenylmethane Triphenylmethane Triphenylmethane Triphenylmethane Triphenylmethane Triphenylmethane Triphenylmethane Triphenylmethane 9-Phenylfluorene 9-Phenylfluorene 9-Phenylfluorene 9-Phenylfluorene a-Saphthyldiphenylmethane Triphenylethylene 1,l-Diphenqlethylene Fluorene Fluorene Fluorene Fluorene Fluorene Fluorene Fluorene F1u or en e Fluorene Fluorene
Cyclohexene concen tiationh
2 2 2 2 2 1 1 1
0 0 0 0
0 9 9 5
1 5
0 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 0
0 0 0 0 0 0 0 0 0 9 9 9 9 5 5 5 5 0
0 0
Hydrocarbon concentration
0 02 10 50 10 10 50 50 2 0 0 01 01 10 50 10 10 50 50 10 10 10 10 50
50
50 50 2 00 2 00
Oxygen pressure'
(Rate R , 2)d
710 300 710 710 710 710 300 710 300 710 710 300 710 710 710 300 710 710 710 500 400 300 710 500 400 300 710 300
0 0472 0476 0394 0254 0123 0252 0194 0100 007 1 0022 0266 0202 0043 0031 0295 0271 0462 0115 0429 0409 0380 0340 0376 03 54 0309 0246 0220 0178
I n chlorobenzene a t 60' in the presence of 0 0504 M A I B S * Concentration in moles/l Pressure in mm and uncorrected for vapor pressure of reaction mixtures Rate in mole/l -hr a
formation of phenol b y an acid-catalyzed decomposition of cumene hydroperoxide. Cobalt bromoacetate in acetic acid appears to generate a low and constant concentration of covalent1s hydrogen bromide. l 9 Dr. H . CoBr(0Ac)
+ HOXc
HBr
+ co(0.X~)~
S. Blanchard has found that cobalt bromoacetate has no effect on the rate of initiation of the AIBN-catalyzed autoxidation of curnene using the Hammond, Sen, and Boozer technique. Pronounced accelerations on the rate of oxidation of cumene (AIBN catalyst) can be duplicated by cobalt(I1) bromide in acetic acid, b u t in this case the initial high rate rapidly decreases due to the formation of an inhibitor, presumably phenol.*" Cobalt(I1) acetate accelerates the AIBN-catalyzed oxidation of cumene but to a lesser extent than the bromoacetate. The effect of cobalt(I1) ion in acetic ( 1 7 ) F F R u i t a n d W E V a u g h a n , I n d Eng Chrm , 41, 259.5 (1919) 118) I hf Kolthoff and A Willman. J A m C/.em. Soc , 66, 1007 (1934). (19) I) A S R a v e n s . Trans I;ai,ada,' S O L 66, , 1768 (19.59) 120) Pi-ivate communication. I)r H S Blanchard. General Electi-ic R e seaich I.abol-,ilory
80
J 0 00
0 08
0 04
Concentration u f triphenylmethane.
Fig. %--Effect of concentration of triplietiylinetll~~~ie on the oxidation of 2.0 ,IT cutiiene in broinobenzene solution :it 60 and 90". In the absence of triphenylmethane a t 60" a n d in the presence of 0.Il504 ,lf A I B S , rate - Ki/% = 0.0144 mole/lLlir, while at 90' in the presence of 0.040 ,\I t-butyl perbenzoate, rate - R , / 2 = 0.00954 mole/l.-lir.
acid may simply be the effect of the cobalt(I1) ion complexing with the peroxy radical R O ~ . C O ( I Ic-) ) RO~:-CO(III)
and affecting the relative values of k , and k t . 2 1 T h e initial rates of oxidation of cumene and cumene -triphenylmethane mixtures in acetic acid in the presence or absence of cobalt(I1) acetate and the bromoacetate are given in Table 111. TABLE I11 RATESOF OXIDATIOX OF CUMESEI N ACETICACID" Concn of (CsHs)CHb
Inhibition, AcceleratorC
Initial r a t e - R , 'pd
7 0
0 Sone 0.0151 0 0 1 Sone 0.0028 80 0.5 Sone Too slow to measure 100 0 CoBr( OAc) 0.160 0 0 1 CoBr( O h ) 0424 15 0.5 CoBr( OAc) ,0014 99 0 CO(OXC)? 0581 0 0 5 Co( OAC)? 0027 96 a 2.0 :If cuniene a t 60", 0.0504 -\f X I B N . MoIes/l. 0.02 M. h'loles/l.-hr.
'
The data of Table 111 show t h a t triphenylmethane inhibits the oxidation of cumene to the same degree in the presence or absence of cobalt(I1) acetate and cobalt bromoacetate. We believe this evidence excludes the possibility that a fast equilibrium between the triphenylmethyl radical and the triphenylmethylperoxy radical is involved.
Discussion The dependence of rate of oxidation of cyclohexene inhibited by polyarylmethanes or polyarylethylenes upon the oxygen pressure clearly indicates that the retardation involves a termination reaction involvinx a radical other than a peroxy radical. The most reasonable interpretation is that this is the polyarylmethyl radical which because of resonance stabilization is able to persist in the presence of oxygen. (21) H S Blanchat-d, .I A m Chori
S Q C ,8 2 , 2014
~l!NiOl
DALEG. HENDRYAND GLENA. RUSSELL
23'74
Bateman has observed that there appears to be an inverse relationship between the relative reactivity of various hydrocarbons toward peroxy radicals and the relative reactivity of the corresponding alkyl radicals toward oxygen. I3 The cooxidations of cyclohexene or cumene ( R H ) with polyarylmethanes (AH) apparently involves the following steps in addition to reactions 1-4.
Vol. SG TABLE IY
CORRELATION
B E T i V E E S I N H I B I T I N G .\BILITY
OF
=\H A S D
THE
STABILITY OF .I. Retarder (.%HI
'3 4.5 1 343 38 1 866 47 1,794 91 1.835 * I n units of d ; ref. 23.
Retardation,"
Fluorene a-Saphthyldiphenyltnethane Triphen>-lmethane 9-Phrn~lfluorene From Table 11, XH = 0.1 -1f.
Calculated delocalization energy of A . b
ethane compared to hexaphenylethaneZ4is due to steric interaction between the substituents rather than due to a greater resonance stabilization of the a-naphthyldiphenylmethyl radical. The greater ability of triphenylmethane to retard the oxidation of cumene a t 90 than a t GOo suggests that a t the higher temperature a relatively greater proportion of triphenylmethyl radicals are present. This may be due to the fact that the reactions in which the nonoxygenated radical disappears have a smaller Ea and thus show less temperature dependence than the reaction whereby the triphenylmethyl radical is formed, ie., hydrogen abstraction by peroxy radicals. A similar phenomenon has been observed in a study of the influence of temperature in the effect of oxygen pressure on the oxidation of unsaturated hydrocarbons. l 3 The retardations observed in the present study are analogous to the inhibiting action of arylethylenes on the oxidation of benza.ldehydes. In particular the retardation of the oxidation of benzaldehyde by Ag.9'bifluorene'j is readily explained by the retardation mechanism demanded by the present results
In the event t h a t the retarder is an olefin (A) such as 1,I-diphenylethylene, step 3a would be replaced by 3a'; R 0 2 . X --t R 0 2 X . . The retardation in rate is thought to result from the fact that k'p, k'', > k , and k , >> k'o. This leads to a situation wherein the rate is reduced (-d[RH]/dt = T k , [ROz.][ R H ] ) because of the reduced concentration of peroxy radicals due to the fact that an appreciable fraction of the total radical concentration is the unreExperimental active A,. In addition, termination by processes 4aApparatus and Procedure.--hutoxidations were performed I d may well occur more readily than reaction 4. using the apparatus and technique described previously.6 Solvents.--.Icetic acid was from freshly opened bottles. The rate expression predicted for steps 1-3d is comCyclohexane, chlorobenzene, and bromobenzene were distilled, plicated. However, it is apparent that if termination chromatographically filtered over silica gel and stored under between a nonoxygenated radical and a peroxy radical nitrogen. is important, the observed retardation should depend Reagents.-Cyclohexene (99.98 minimum mole c > )and cumene (100.0 minimum mole r; were obtained from Phillips Petroleum on : ( 1 ) polyarylmethane concentration, (2) the rate Co. Tlie AIBS used was described previously.6 Triplienylconstant for attack of peroxy radicals on the polyarylmethane and fluorene were obtained from Matheson Coleman and methane, and (3) the oxygen concentration. In refBell while triphenylethylene and 1,l-diphenylethylene were from erence to item (2) it has been mentioned that the ability r\ldrich Chemical Co. 9-Phenylfluorene from an unknown of a hydrocarbon to retard the oxidation of cyclohexsource was recrystallized; m . p . 145-147'. a-Xaphthyldiphcnylmethane was prepared by reduction of the carbinol obtained by ene is qualitatively proportional to the expected stat h e addition of u~tiaphttiylrnagnesium bromide to benzophcnone.2fi bility of the resulting polyarylmethyl radical. Table Cobalt bromoacetate was prepared by the addition of 1 equivalent IT' demonstrates that this prediction is fulfilled, but of anhydrous hydrogen bromide t o anhydrous cobalt( I1 I acetate that cu-naphthyldiphenylmethane is not as efficient an in acetic acid solution. Tlie hydrogen bromide \vas nieasuretl i n a calibrated standard bulb of a Stock-type high vacuum apinhibitor as the calculated delocalization energy of the paratus. radical predicts. This is presumably due to steric inhibition of resonance which is well recognized for the (23) G W Wheland, J Chem P h y r , 2 , 47-1 8,1934) (24) G W Wheland, "Advanced Organic Chemistry," :3rd Ed . Joh.1 corresponding carbonium ion. 2 2 \Vliley and Sons, I n c , N e w York, N.Y .IgBO, p 7 7 1 The present results suggest that the higher degree of (231 G Wittig and \V Lange. A n n , 636, 266 (19:38), G WittiE a n d G Pieper, i b i i i , 646, 1-12, 172 11441): 658, 207, 218 (11147). G. Wittig, dissociation of 1.2-cIi-cu-naphthyl-1,1,2,2-tetraphenyl-
+
1221 A Streitwieser,
JI
, J . A I ! ? .C h r i n . SOL, 7 4 , 52% (1952)
h b i d . , 658, 201 (19-15) ( 2 6 ) S. F. Acree. APY , 37, 616, 6 2 5 (1904).
