ORGANIC PEROXIDES SOME PROPERTIES OF ORGANIC PEROXIDES BY A. C. EGERTON, W. EMTE AND G. J. MINKOFF Received 22nd February, 1951 A number of alkyl bydro-, di-alkyl and acyl hydrb-peroxides have been prepared, and several of their physical properties have been studied. The refractive indices and melting points are recorded and compared with values in the literature. The vapour pressures have been measured in a static system, and the latent heats of vaporization deduced from the data. The heats of com- bustion of the peroxides were measured in a bomb calorimeter, leading t o values of the heats of formation. Comparisons are made between the heats of formation of R . 00 . R' and of R . 0 . R' ; in general, differences in heats of formation are 35-50 kcal., though acyl hydro-peroxides give results more difficult to interpret.Investigations have been made in this laboratory relating to the slow, and rapid, combustion of hydrocarbons, e.g. by Bone and Hil1,l Newitt and Thornes,S G a y d ~ n , ~ Egerton and Minkoff,4 and Egerton, Harris and Young.6 That the peroxides play a role of importance in combustion, particularly in relation to the phenomenon of " knock ID, has long been recognized (Egerton, Callendar (1927)). Investigation of their decom- position was therefore made by Harris and Egerton.6 I n spite of these investigations and many others elsewhere, there is still doubt as to the particular peroxide which is formed under different circumstances during hydrocarbon combustion. The object of the present investigation was to gain familiarity with some of the different types of simpler peroxides, and to determine some of their properties.The following have been investigated : alkyl hydroperoxides, dialkyl peroxides, peracids, monohydroxy-alkylhydroperoxides, dihydroxy-dial- kyl peroxides and monohydroxy dialkyl pkroxides. Methods of preparation have been described by Baeyer and Villiger,' by Rieche and Hitz 8 and by D'Ans and Frey ; they have been further applied by Harris and Egerton,6 Harris,lO Milas and Surgenor and particularly by Eggersgluss and Schutt (1941). The latter work was Bone and Hill, Proc. Roy. SOC. A , 1930, 129, 434. Gaydon, Proc. Roy. SOC. A , 1942, 179, 435. Egerton and Minkoff, Proc. Roy. SOC. A , 1947, 191, 145. a Newitt and Thornes, J. Ckem. SOC., 1937, 1656. 6 Egerton, Harris and Young, Trans.Faraday SOC., 1948, 44, 745. 6 Egerton and Harris, PYOC. Roy. SOC. A , 1938, 168, I. 7 Baeyer and Villiger, Bey. B, 190. 33, 2479, etc. 8 Rieche and Hitz, Ber. B, 62, 2458, etc. D'Ans and Frey, Ber. B, 1912, 45, 1848. loHarris, Proc. Roy. SOC. A , 1939, 173, 126. 11 Milas and Suigenor, J Amer. Chem. SOC., 1946, 68, 205.A. C. EGERTON, W. EMTE AND G. J. MINKOFF 279 not published, but an account of i t was available and one of the present authors (W. E.) took part in the investigation. The methods of prepara- tion will not be described in detail in this communication. The preparation of the peroxides in a state of purity is still difficult. The properties which have been measured are the vapour pressures, refractivities, melting points and heats of combustion. The properties of hydroxy-peroxides will be considered in a later paper. Experimental The Vapour Pressures, Melting Points and Refractivities .-The static method for measurement of vapour pressure was chosen.It was possible t o use a differential manometer containing mercury for di-tert.-butyl peroxide, which is comparatively stable, and even, with certain precautions, for diethyl- peroxide, but most peroxides attack mercury and a spoon gauge F (see Fig. I ) , was used with all the peroxides except di-terl.-butyl peroxide. The peroxide was placed in B and frozen with liquid air : the apparatus was then evacuated. After closing tap TI, the peroxide was distilled from B t o C and back from C to B, the middle fractions being rltained and the rest pumped off.Finally the peroxide was distilled into A and, after evacuating the spoon gauge with A cooled in liquid air, the apparatus sealed off a t D. Counter pressure was pro- vided from nitrogen in bulb E to bring the pointer back t o zero (using a mirror in conjunction with a Pointolite lamp) : 4 scale divisions were equivalent to I mm. Hg pressure difference. The vapour pIessure could be measured on the manometer to 0-1 mm. Hg. Observations were made a t six t o ten different temperatures, the mean of five a t each temperature being plotted ; the separate determinations agreed to about I yo. Up t o room temperature A was main- tained a t the required temperature by immersing the container in cold alcohol and above room temperatlie in water, both baths being weil stirred.The calibrated mercury thermometers could be read t o 0.01' C. Leakage or de- composition could be detected by freezing the peroxide and measuring the residual pressure : measurements were discarded when the presence of permanent gases was indicated. Materials .-METHYLHYDROPEROXIDE was made from dimethylsulphate and hydrogen peroxide (30 yo) ; the distillate was extracted with ether and frac- tionated.2 80 ORGANIC PEROXIDES ETHYL HYDROPEROXIDE was prepared similarly from diethylsulphate. i!ert.-BuTYL HYDROPEROXIDE was obtained by the action of 85 yo H,O, on tert.-butyl alcohol in presence of anhydrous magnesium sulphate. After remcval of alcohol, the residue was fractionated and then redistilled. DIETHYLPEROXIDE was prepared from diethyl sulpbate and 30 yo hydrogen peroxide ; the oily layer was separared from that containing most of the ethyl hydrogen peroxide, distilled and fractionated.DI-fert.-BUTYL PEROXIDE was prepared from butyl hydrogen sulphate (ob- tained via tert.-butyl alcohol) and 30 yo hydrogen peroxide. The oily layer was fractionated and redistilled. PERACIDS (acetyl hydroperoxide, propionyl hydroperoxide and butyryl- hydroperoxide) were prepared by perhydrolysis of the diboro-carboxylic acids, followed by fractionation. Some samples were also prepared from the acid anhydrides and hydrogen peroxide, but the product was generally less pure by this method. The impurities were a small peicentage of tbe corresponding acid and occasionally some ethyl ether. HYDROGEN I?EROXIDE.-T~~S was a redistillate cf a IOO yo H,O, obtained from Messrs.Laporte. Results The results of the vapour pressure measurements were plotted as log p against I/T and the equations of the straight lines obtained are given in Table I, together with the value of the latent heat of vaporization within the range of the measurements. TABLE I.-VAPOUR PRESSURES Peroxide Methyl hydroperoxide . Ethyl hydroperoxide . tart.-Butyl hydroperoxide . Diethyl peroxide . Di-tart.-butyl peroxide . Acetyl hydroperoxide . Propionyl hydroperoxide . Butyryl hydroperoxide . Hydrogen peroxide . iange of Measurements OC - 20 t o + 40 - 20 t o + go 0 to I20 - 20 to + 60 - 20 t o + I00 0 t o I10 0 to I20 0 to I20 o t o 60 L(ca1.) 9020 10190 11130 6930 7420 105 70 10320 10860 I I 300 8.38 -1972/T 8.834-2228JT 8.891-2432JT 7-140-1621 JT 8-91 I - 23 I I/T 8.623 - 2256 JT 7.356- 1517/T 8.83 -2376/T 8-63 -2469/T The slopes of the lines are approximately the same except for those of the dialkyl peroxides which are less steep.The Trouton constant (LIT,) is about normal for the dialkyl peroxides, but is high for the others. At the higher temperatures, above the range over which the equation holds, some decomposi- tion was observed with methyl hydroperoxide and hydrogen peroxide, and t o a less extent with ethyl hydroperoxide. The refractivities of these peroxides determined with a Hilger-AbbB refracto- meter are recorded in Table 11. The values in column 4 were obtained, using different preparations, subsequent to the vapour pressure determinations by one of the authors (G.J. M.). The values given in column 5 are quoted from the literature. The melting points were also determined with a thermometer (Reichenstalt calibration) and a thermocouple. Considering the difficulty of obtaining the peroxides in a pure state, there is fair agreement between the values. The freezing point of methyl hydroperoxide is higher than that given by Rieche. The vapour pressure of this peroxide also appears t o be considerably @eater than the value of 65 mm. at 3g0 C quoted by the same author. The vapour pressure of 2ert.-butyl hydroperoxide is rather lower than 15 mm. a t 37" C quoted elsewhere. The peracids are particularly difficult to obtain in a pure state and there is some doubt as to the accuracy of the values given, although they agree reasonably well with other determinations.The acetyl hydroperoxide was made by two quite different methods, and the refractivities agreed to 0.2 yo ; nevertheless, the acetic acid content was different. (Iodometry is of little use for estimation of the purity of the peroxides.) That the slopes or the vapour pressure lines are of the expected magnitude is shownA. C. EGERTON, W. EMTE AND G. J. MINKOFF 281 by comparison with those given for (a) ethyl alcohol, (b) diethyl ether and (c) tert.-butyl alcohol, viz., (a) log P = 9.205 - 2216/T ; L = 10140 cal./mole. (b) log P = 7-92 - 1543/T ; L = 7059 cal./mole. (c) log P = 9-53 - 2361/T ; L = 10800 cal./mole. TABLE II.-REFRACTIVITIES Peroxide Methyl hydroperoxide . Ethyl hydroperoxide . tert.-butyl hydroperoxide .Diethyl peroxide . Di-tert.-butyl peroxide . Acetyl hydroperoxide . Propionyl hydroperoxide . Butyryl hydroperoxide . Hydrogen peroxide . Peroxide ,I 5 O D 1.3646 1.3832 1.4027 1.3712 1.3908 1'3994 1.4101 1.4148 1.4086 ,209 D 1.3608 1.3801 1.4007 1.3888 1'3974 1.4041 1'4125 1.4058 - R = determination by Reiche. TABLE III.-MELTING POINTS ,200 D 1.3648 I 1'4009 1.3698 1'3905 1.4022 1.4057 - - Methyl hydroperoxide Ethyl hydroperoxide . tert.-Butyl hydroperoxide Diethyl peroxide. . Di-tert. -butyl per oxide Acetyl hydroperoxide Propionyl hydroperoxide Butyryl hydroperoxide Hydrogen peroxide . Melting Points C -72 to -78 - 8 t o - I 0 - 68 to - 69.5 - 18 t o - 19'5 glass 0'2 - - 13-1 - 10'0 - 1.8 Literature Value Literature Value OC - I00 t o 100.5 - I00 t o 100.5 - 13'5 - 70 - 18 - 13'3 - 10.3 - 1.7 0'1 The latent heat change due to the introduction of the extra oxygen atom in the molecule is quite small, but the vapour pressures of the peroxides are, however, lower, except for the peracids where the position is reversed.The boiling points in O C estimated by extrapolating the plots t o log p = 2.8808 are as follows : Methyl hydroperoxide . . 86 Methyl alcohol . (64'5) Ethyl hydroperoxide . 95 Ethyl alcohol . - 77 (78.3) Diethyl peroxide . . 66 Diethyl ether . * 33'5 (34'6) terl.-Butyl hydroperoxide . 133 ten!.-Butyl alcohol . 82 (82.5) Di-tert.-butyl peroxide . . 108 Di-tert.-butyl ether . Acetyl hydroperoxide . . IIO Acetic acid . . 118.5 Propionyl hydroperoxide . 120 Propionic acid . . 140 Butyryl hydroperoxide .. 126 Butyric acid . . 162-3 Hydrogen peroxide . - I57 The figures in parentheses indicate the true b.p (literature values). The values obtained from measurements of vapour pressures and extrapolation would be expected io give somewhat lower values than the true boiling points, but they are probably not more than about a degree lower. Decomposition prevents their accurate determination. The value obtained for diethyl per- oxide by Rieche, 64-5O a t 760 mm., is lower than the above estimate. The value found by direct determination of the comparatively stable di-tevt.-butyl peroxide was I O ~ O C .282 OXIDATION OF TRIMETHYLETHYLENE Heats of Combustion.-A number of determinations of the heats of combus- tion of the above mentioned peroxides have been made during the course of this work by the bomb calorimeter method and the authors are indebted to the Director of the Fuel Research Station for permission to use the apparatus a t the Station for the purpose. The results, however, show discrepancies, presumably due to the difficulty of ensuring complete combustion, and further determinations are in hand. The difference between the heats of formation of water vapour and of hydrogen peroxide vapour is 24-2 kcal. The difference, however, between the heats of formation from the atoms, of hydrogen peroxide and of water vapour is 34.8 kcal./mole. This is the energy difference due to the -0-0- group instead of the -0- in the molecule and t o the consequent change in the strength of the OH bands. One of us (G. J. M.) wishes to express his gratitude to the University of London for the award of an I.C.I. Research Fellowship. We are grateful to the Director of the Fuel Research Station, Greenwich, for permission to use the bomb calorimeter. Department of Chemical Engineering and Applied Physical Chemistry, Imperial College, s. w.7.