Jan., 1958
RADIATION OF n-HEXANE
the 3975 A. peak disappeared at this temperature. I n order to determine whether radiolysis produced the same spectral changes as photolysis, tritiated ethyl iodide with a specific activity of 0.5 curies per milliliter was allowed to stand as the glass a t -190". Its spectrum could not be taken before melting since the self-irradiated glass always cracked after about one-half hour. When, after 41 hours as the glass, it was melted gnd refrozen, its spectrum showed a peak a t 3975 A. with a maximum absorbance of 0.5. The nature of the species causing these peaks is not fully known. However, when hydrogen iodide was added to a cell containhg pure ethyl iodide, a peak was observed a t 3700 A. exhibiting identical behavior to that produced by the photolysis of pure ethyl iodide. The peak a t 3975 A. was observed when both hydrogen iodide and iodine were added to ethyl iodide. Neither of these peaks appear in the spectrum of the liquid a t room temperature. It appears, therefore, that a complex between hydrogen iodide and ethyl iodide and one between hydrogen iodide, iodine and ethyl iodide are formed a t low temperatures. The fact that this complex does not appear in the spectrum of the glass after the photolysis of ethyl iodide, until the glass has been melted and refrozen, suggests that either hydrogen iodide is not produced in molecular form in the glass, but is formed when the glass is melted, or that the activation energy for the formation of the complex is high enough to prevent its formation a t - 190".
15
Other Comparisons of Liquid and Solid Phase Yields.-Norman and Porter4b have produced a number of "frozen in" free radicals by the photolysis of the parent compounds in rigid glasses of 3methylpentane and also of ether-isopentane-ethanol-(EPA). I n the course of these studies they have observed that the quantum yield of iodine production from ethyl iodide, by 2537 8. light in the solid glasses a t - 196" is 0.5 of that in the liquid a t 20". They report that the iodine color was not visible until the glass had been warmed. This observation suggests the presence of "trapped" atoms. The probability of dissociation in rigid media seems to depend on the size of the radical fragments. For instance, a spectrum assigned to the benzyl radical has been reported4b to have been produced in the photolysis of toluene, benzyl chloride, benzyl alcohol and benzylamine in EPA and hydrocarbon glasses. This spectrum was not produced, however, in similar solutions of diphenylmethane or dibenzyl. I n addition, acetophenone and diphenylmercury have a negligible quantum yield for photodissociation in isopentane-3-methylpentane glasses compared to the liquid solution.4c Additional references to related work are given in the references3+ cited above. Acknowledgments.-This work was supported in part by the University Research Committee with funds made available by the Wisconsin Alumni Research Foundation and in part by the Atomic Energy Commission. It also was aided by a fellowship provided by the Standard Oil Foundation.
RADIATION CHEMISTRY OF ORGANIC COMPOUNDS. 11. n-HEXANE BY H. A. DEWHURST General Electric Research Laboratory, Schenectady, N . Y . Received June 84, 1867
The formation of gas and liquid products in the 800 kvp. electron radiolysis of n-hexane liquid has been studied as a function of the energy absorbed, temperature and the presence of solutes. The hydrogen yield decreased with tem erature to a limiting value in the solid state. The same limiting yiela was attained in the resence of solutes. The yield ofunsaturated in some cases, to decrease in the presence of product, mainly trans-vinylene wa6 independent of temperature but was solutes, notably oxygen and iodine. The yield of intermediate and dimer product was found t o depend on temperature as well as the solute present. The results are discussed in terms of ion-molecule, radical and so-called molecular processes.
faun%,
1. Introduction The nature of the products formed in the radiolysis of n-alkane liquids was reported in a previous paper.' The diversity of product formation indicated that extensive C-H and C-C bond rupture had occurred. Apart from end effects and physical state the small hydrocarbon molecules are useful as model compounds for long chain hydrocarbon polymers such as polyethylene. A better understanding of the formation of non-volatile products in model compounds could be useful for the elucidation of processes which occur in polymers. Previous studies2 with n-hexane were concerned mainly with the gaseous products. Henri and co(1) H.A. Dewhurst, A.C.S. Miami Meeting, April 1957; THISJOUR61, 1466 (1957). (2) C. S. Sahoepqe and C. H. Fellows, Ind. Eng. Chem., 23, 1396 (1931).
NAL,
workers3 examined the a-particle decomposition of n-hexane in the gas phase. The authors worked at very high conversions (40%) and found that 73% of the reacted n-hexane was in a non-volatile liquid. Recently Krenz4 has studied the y-ray decomposition of n-hexane liquid. He showed that anthracene M ) had no measurable effect upon the gas yields but caused a diminution in the yield of unsaturated products. The radiolysis of n-hexane liquid was examined in order to evaluate the processes responsible for product formation. Gas-liquid partition chromatography was used to examine the formation of intermediate and dimer product under a variety of experimental conditions. (3) V. P. Henri, C. R . Maxwell, W. C. White and D. C. Peterson, THra JOURNAL, 66, 153 (1952). (4) F. H. Krenz, Nature, 116, 1113 (1955).
H. A. DEWHURST
16
T
B.F!
I
I
-150
-100
I
I
-50
0
TEMPERATURE,
Fig. 1.-Effect
I
I
50
100
OC.
of temperature on the hydrogen yield; energy absorbed, 6 X lozoe.v.
2. Experimental The methods employed were similar to those described in a previous paper.’ 2.1 Materials.-The n-hexane used in these studies was Phillips pure grade. As received the hexane contained a small amount of ultraviolet absorbing impurities which were removed readily by a single silica gel treatment. The silica gel treated hexane was found by gas-liquid partition chromatography to contain small amounts of n-pentane and hexane isomers. Washing with sulfuric acid followed by fractional distillation removed these impurities.5 It was found that the hydrogen yield and unsaturation yield was not affected by the purification methods. The following solutes were used: p-benzoquinone (resublimed), t-butyl anthraquinone (recrystallized), cyclohexene (Phillips pure grade), benzene (Phillips research grade) and 1 2 (Mallinckrodt). 2.2 Irradiation.-The samples were irradiated with 800 kvp. electrons from a resonant transformer unit.e Thin window glass cells were used for the gas studies. The samples were thoroughly de assed by freeze-pump technique. The hydrogen peld, G(H5 = 5.3, from cyclohexane was used as a chemcal dosimeter.’ For low temperature irradiation the glass cells were placed in an aluminum jacket and irradiated in a small Dewar flask with the appropriate freezing mixture. T o obtain high radiation doses in a reasonable tim; samples were irradiated in a thin la er (- 1 mm.) in a 2 diameter aluminum cell from whici oxygen could be excluded. The cell was mounted on a thermostated metal block a fixed distance from the generator window. The amount of energy absorbed by the liquid was determined with an ionization chamber of identical geometry.* 2.3 Product Analyses.-Conventional high vacuum gas analysis techniques were employed. The amount of unsaturation was determined by infrared analysis usin a Perkin-Elmer IR spectrometer (Model No. 21). A f’erkinElmer vapor fractometer (Model 154) was used for the gas chromatography analysis. The methods employed have been described .I
3. Results and Discussion Hydrogen Formation.-It has been thoroughly established that the volatile products from n-alkane liquids consist largely of hydrogen with smaller amounts of saturated and unsaturated hydrocarbons. 3.1.1 Effect of Temperature,-The effect of temperature on hydrogen formation is shown in Fig. 1. In the liquid phase between -78 and 30°, the yield increased approximately 3% for every 10” 3.1
(5) The author is indebted to Mr. E. B. Cox for several samples of purified hexane. (6) E. E. Charlton and W. F. Westendrop, Q. E. Rsr., 44, 652 (1941). (7) R. H. Schuler and A. 0. Allen, J . A m . Cham. Hoc., 7 7 , 507 (1955). (8) E. J. Lawton and J. 5. Balwit, private aommunication.
Vol. 62
increase in temperature. This small temperature coefficient corresponds to an over-all activation energy of about 3 kcal./mole. The significance of this temperature coefficient is uncertain; however, it is of interest to note that it is about the same as the temperature coefficient for diffusion processe~.~ This correlation may reflect an effect of temperature on the diffusion expansion of a spur. I n the solid phase, between liquid nitrogen temperature and - 120°, the hydrogen yield appears to be independent of the temperature. These samples were analyzed after bringing the cell to room temperature. The contribution of H-atom abstraction to the hydrogen yield would be expected to be negligible a t these low temperatures unless the abstraction involves hot H-atoms. The hydrogen yield in the solid state could arise from molecular processes and/or H-atom recombination. The scavenger experiments described below indicate that the hydrogen from the solid state may be entirely molecular in origin. The definition of a molecular process is uncertain; it could represent hot atom processes which would be temperature independent. 3.1.2 Effect of Solutes.-The effect of a variety of scavenger molecules on the hydrogen yield is shown in Fig. 2. These studies were made at room temperature. The hydrogen yield decreases with increase in scavenger concentration and attains a limiting value which is approximately the same for the scavengers studied. The highest scavenger concentration shown in Fig. 2 is 0.038 M . Further studies with cyclohexene at concentrations of 0.18 and 0.50 M showed that the same limiting value of the hydrogen yield was attained. It is of considerable interest that the scavenger-limited hydrogen yield is the same as the hydrogen yield in the solid state a t low temperatures. These results show that 60% of the hydrogen yield a t 30’ is formed by processes which are temperature and scavenger independent and have therefore been attributed to molecular processes. Dorfman’O has shown that in the radiolysis of ethane a large fraction of the hydrogen is formed intramolecularly without the intermediate formation of single atoms of hydrogen. Schuler’l has found that 1 2 a t conce’ntrations up to 0.20 M in benzene has only a small effect on the hydrogen yield. With cyclohexane he found that I1 decreased the hydrogen yield to a limiting value which represented 50% of the total hydrogen. The mechanism of the molecular detachment process is not completely understood. There is evidence from mass spectral fragmentation patterns which shows that the loss of 2 H-atoms is an important process for the saturated hydrocarbons. The loss of 2 H atoms as a molecule of Hz would be a lower energy process and would presumably correspond to molecular hydrogen detachment. Further discussion of the mechanism is given in section 3.2. 3.2 Unsaturation.-It has been shown’ by infrared analysis that the unsaturation produced in n-hexane is mainly of the trans-vinylene type with a small amount of vinyl type. Gas chromatography (9) 5. Glasstone, I(.J. Laidler and H. Eyring, “Theory of Rate Procemes,” MoGraw-Hill Book Co., Ino.. New York. N. Y. (10) L. hl. Dorfman, THIS JOURNAL,60, 826 (1956). (11) R. H. Sohuler, ibid., 60, 381 (1950).
4
RADIATION OF ?&-HEXANE
Jan., 1958
has shown that part of the trans-olefin is hexene-2. The formation of double bonds as a function of the energy absorbed is shown in Fig. 3. The gradual decrease in the trans unsaturation yield has been attributed to scavenger action of the olefin product. From the initial slopes of the curves in Fig. 3 the following initial yields were calculated
17
0
-
b
u
G(trans) 1.2 G(viny1) = 0.3
0 t . B U T I L ANTHRAPUINONE p.BENZO OUINONE
A IODINE so that G (unsaturation) = 1.5. Krenz4 has reported that G(unsaturation) = 4 f 2. The reason for this discrepancy is not apparent; however, the large error limits imposed by Krena will not allow SCAVENGER CONCENTRPTION, MOLES/LITER a rigorous comparison to be made. It is of inter- Fig. 2.-Effect of scavengers on the hydrogen yield: energy absorbed, 6 X lozoe.v. est to note that in the presence of anthracene, he found that G (unsaturation) approached 2. 3.2.1 Effect of Temperature.-The effect of 3.5 0 0 30’C temperature on the unsaturation yield was exLIQUID NITROGEN amined a t liquid nitrogen temperature and room temperature. The results in Fig. 3 show that the 3.00 unsaturation yield is independent of temperature. These results suggest that the unsaturated prod- *. 0 ucts are not formed by radical disproportionation x 2.5reactions but rather by molecular processes. Such \l3 processes would yield an equivalent amount of hy- mW drogen and the over-all reaction can be written as d 2.0 C6&4 - w d C B I S HZ bThe hydrogen formed in this process represents ap- 3 0 1,5 proximately one-half the molecular hydrogen yield K discussed in section 3.1. 3.2.2 Effect of Solutes.-In view of the tem1.0 V I N/ Y OL perature independence for G(unsaturation), it was of interest to examine the effect of scavengers. The results in Table I show that some scavengers decrease the trans unsaturation yield while others have little or no effect. X IO’.
s
+
I
TABLE I EFFECT OF SCAVENQERS ON THE FORMATION OF DOUBLE
BONDS Energy adsorbed, Q(molecuIea/100 e.v./g. e.v.1 X IO-*’ trans vinyl
0
5
I
10 ENERGY
I
I
1
15 20 25 ABSORBED, eV/G X
I
30
Fig. 3.-Formation of unsaturation as a function of energy absorbed and temperature.
carbonyl absorption can be due to direct radical addition to the carbonyl as well as to the ring.12 None .. 5.8 1.0 0.30 The decrease in trans yield in the presence of tCyclohexene 0.18 8.7 0.85 .35 butylanthraquinone is uncertain since the quinone Cyclohexene .50 11.6 .93 .25 has a weak absorption band in the region of trans Benzene .45 5.8 .80 .4 absorption, 10.35 p . The solutions remained clear p-Benzoquinone .038 2.9 1.3 .. throughout the irradiation and in fact became more 5.8 0.80 .27 intensely colored. At low doses the carbonyl and l-Butylanthraquihone .028 Satd. 2.9 .20 .O Oxygen aromatic double bond absorption bands were un2.9 .60 .O changed, whereas the absorption band attributed to Iodine -0.03 the t-butyl group decreased. Cyclohexene and benzene appear to have a small The yield of trans and vinyl double bonds was effect on the trans yield. With cyclohexene high markedly decreased in the presence of oxygen. Indoses were employed and consequently one would frared analysis showed that hydroxyl, carbonyl expect a small decrease (cf. Fig. 3). With ben- and ether bonds had been formed. Iodometric tizene, however, the effect appears to be real and is tration showed that both dialkyl peroxides and alin the direction which suggests that energy trans- kyl hydroperoxides also were formed.13 fer protection may be operative. 3.3 Gas Chromatography.-In part I’ it was In the presence of p-benzoquinone the trans un- shown that the radiolysis products from n-hexane saturation yield was unaffected; however, a copi- could be classified into essentially three groups, ous green-black precipitate formed which indi- namely, low molecular weight products (C, to cated the formation of a quinhydrone. Infra- C$, intermediate molecular weight products (Cy red analysis of the filtered solution showed that the (12) Of. P. J. Flory. “Prineiplea of Polymer Chemiatry,” Cornell carbonyl absorption of p-benzoquinone disappeared University Presa, Ithaca, N. Y.. 1953, p. 184. very rapidly on irradiation. The disappearance of (13) L. E. St. Pierre and H. A. Dewhuret, unpublished results. Scavenger
Concn. molea/l.
i$
B. A. DEWHURST
I
6
/
I
/b
k
0
ENERGY ABSORBED, e v / p
X
Fig. 4.-Product formation, determined by gas chromatography as a function of the energy absorbed.
to CI,) and dimers (GI,) hereafter referred t o as C1, Ci and c d , respectively. The formation of each product group was studied as a function of the energy absorbed and the results are shown in Fig. 4. The product concentration is expressed as relative peak area per cent. and was determined by peak area integration with a planimeter. I n separate experiments it was shown that the relative peak area was directly proportional to the weight per cent. of the product. The results for the intermediate and dimer product groups were obtained with thecolumn at 150" while the results for the low molecular weight group were obtained with the column a t 50". The data in Fig. 4 show that product formation is directly proportional to the energy absorbed up to 9 X loz1e.v./g. which indicates that these products are in fact primary products and do not result from further reactions of olefins or other radiolysis products. The initial slopes of these curves were used to calculate initial yields G(Ci2) = 2.0 G(Ci ) = 1.5 G(C1) = 2.114
The experimeiital error in these values was estimated as 10%. 3.3.1 Product Distribution.-The product distribution in the low molecular weight group was discussed in part I. It was shown that in this group each product peak on the chromatogram consisted of saturated and unsaturated hydrocarbon. These products could be formed by carboncarbon bond scission followed by disproportionation reactions. There is evidence from scavenger studies that these products may be formed by proc(14) T h e results for CI in Fig. 4 are low b y almost a factor of three due t o evaporation loss. A complete CI product analysis gave: G(CHd = 0.15,G(CzHd = 0.30,G(CzHe) 0.30, G(CaH6) = 0.13, 0.20. G(CsH8) = 0.42,G(C4) = 0.53and G(Cs)
-
Vol. 62
esses which do not involve free radical intermediates. A complete identification of the products in the intermediate and dimer groups was not made. Retention time studies and mass spectrometer identification of isolated products showed that the Ci and C d products were mainly saturated branched chain hydrocarbons. Unsaturated products were not found. The nature and diversity of product formation indicates the importance of both C-C and C-H bond rupture processes. If it is assumed that the C7 to CII products are formed by combination reactions of radicals from C-C bond rupture with radicals from C-H bond rupture, hexyl radicals, and that the Clz products are formed by combination of hexyl radicals, then these several products would be expected. C,: n-heptane 2-methylhexane 3-methylhexane Cg: n-nonane 4-methyloctane 4-ethylheptane '211: n-undecane 5-methyldecane 4-ethy lnonane
i
V
CS: n-octane 3-methylheptane 3-ethylhexane Cl0: n-decane 5-methylnonane 4ethyloctane Clz: n-dodecane 5-methylundecane 4ethyldecane 5,6-dimethyldecane 4ethyl-5-methylnonane 4,5-diethyloctane
The distribution of product peaks on the chromatogram is in qualitative agreement with the above expected products. Furthermore, mass spectrometric analysis of the isolated Csfraction has shown the formation of the expected octane isomers. Thus the general trend of product formation has been established. Assuming that the C i and c d products are formed by radical combination reactions, the following radical yields can be calculated G(Ce radical) = 5.5 G(Ci radicals) = 1.5 where i = 2, 3 and 4
3.3.2 Effect of Temperature.-The chromatogram of hexane irradiated a t liquid nitrogen temperature showed that the Ci product was almost completely eliminated and that the C d product was decreased by one-half. These results are consistent with the suggestion that C i and c d products are formed by radical combination. reactions and further indicate that part of the Ca product may be molecular in origin. However it has been shown (Sec. 3.3.3) that in the presence of iodine or oxygen the dimer products are almost completely eliminated. Therefore it is proposed that the dimer formation at liquid nitrogen temperature occurs by combination of hexyl radicals formed by the reaction C6Hll+
+ C&4 + e- *2CeH13 + Ha
It is noteworthy that this reaction not only gives hexyl radicals but also an equivalent amount of molecular hydrogen. 3.3.3 Effect of Solutes.-It was found that radical scavengers also decreased the amount of C i and Cd product. The efficiency of radical trapping as measured by the decrease in product formation was found t o depend on the scavenger mole-
,
Jan., 1958
RADIATION OF %HEXANE
cule. Oxygen and iodine were found to be the most efficient scavengers since the Ci and Ca products were almost completely suppressed. Some typical results are summarized in Table 11. TABLE I1 EFFECT OF SOLUTES O N RADIOLYSIS OF +HEXANE Solute
None None Cyclohexene (0.18 M) Benzene (0.45 M) Ethylene Oxygen Iodine (0.03 M )
Dose, e.v./g.
x
10-21
2.9 8.7 2.9 8.7 8.7 8.7 2.9 2.9
CI
Product ooncn. (area %)
.. 0.81
..
.84 .82
..
..
..
Ci
C1n
0.75 2.9 0.40 1.6 1.3 2.0 0.0 0.0
1.3 4.3 0.60 2.3 2.3 3.7 0.0 0.0
In the presence of cyclohexene new product peaks were observed on the chromatogram. One of these peaks was identified as cyclohexane while the others corresponded to higher boiling addition products. From the peak areas it was found that the average decrease in cyclohexene concentration corresponded to a net yield of G(-CeHlo) = 2.1, while the yield for cyclohexane formation was G(C6H12) = 0.8. Under the same experimental conditions it was found that the decrease in hydrogen yield was G(-AH2) = 2.0 and thus the disappearance of one cyclohexene molecule corresponds to the removal of one hydrogen molecule. The appreciable yield of cyclohexane is of special interest since it may indicate the existence of regions of high atom concentration. The high efficiency of iodine as a radical scavenger has been utilized by Hamill and otherslS-17 to study free radical formation in the radiolysis of hydrocarbons. A sample of hexane saturated with I2(-0.03 M ) was irradiated to a dose of 2.9 X 1021e.v./g. The chromatogram obtained a t 125" showed that small amounts of ethyl, propyl and butyl iodides were formed, pentyl iodide was just detectable. Two large product peaks a t 19.3 and 26.5 minutes corresponded to sec-hexyl iodide and n-hexyl iodide, respectively. The 19.3 minute product was isolated in a liquid nitrogen trap and its identification confirmed by mass spectrometry. Chromatograms obtained a t lower column temperatures showed the formation of only small amounts of methyl iodide. From the peak areas the following yields were calculated Iodide
G(RU
Methyl Ethyl n-Propyl n-Butyl sec-Hexyl n-Hexyl
N O ,10
19
The G-value for total mono-iodide formation is very close to that recently reported by McCauley and Schuler18for liquid n-butane. If the iodide products are assumed to represent the distribution of radical species then a direct comparison with the distribution of products in absence of iodine should be possible. In the presence of iodine these radical yields were calculated G(Ce radicals) = 3.3 G(Ci radicals) = 1.5
It is seen that the Ci radical yield is in good agreement with that obtained from the product distribution, see. 3.3.1, but that the C6 radical yield is appreciably lower in the presence of iodine. The decreased c6 radical yield can be explained by the competitive reactions H C e b +Cs& 4- H2
+H +
I2
+HI + I
which is also consistent with the observed decrease in hydrogen yield in the presence of iodine. As shown in Table I1 oxygen was found to be a very efficient inhibitor for the Ci and Cd products. The chromatogram obtained a t 150" of hexane irradiated in the presence of oxygen showed that most of the product was contained in a poorly resolved triplet which had the same retention time as c 6 carbonyl. This triplet was resolved at lower column temperature and the products identified by retention time and by mass spectrometry of the isolated product. The product was found t o be mainly hexanone-2 and hexanone-3. The chromatogram clearly shows that in the presence of oxygen the intermediate and dimer products are almost completely eliminated. Bachlg has reported that in the presence of oxygen heptane formed dialkyl peroxide and smaller amounts of hydroperoxide and hydrogen peroxide. We have observed similar products by iodometric titration. l3 The peroxidic products presumably would have decomposed rapidly in the fractometer column a t 150". From the integrated peak areas the total yield for c6 carbonyl product was G = 6.0 in good agreement with the data of Bach for total oxidation products. The results indicate a nonchain oxidation mechanism from which the c6 carbonyl yield may be a direct measure of the hexyl radical production. 4. Conclusions From the combined studies by gas analysis, infrared analysis and gas chromatography the hydrogen yield from the radiolysis of n-hexane can be summarized as follows
5t
t
.40 .70 .30 2.60 0.70
(15) W. H. Hamill and R. R. Williams, Jr., J. A m . Chem. Soe., 72, 1857 (1050). (16) L. H. Gevantman and R. R. Williaiiis, Jr., THISJOURNAL, S6, 5G9 (1952). (17) E. N. Weber, P. F. Forsyth and R. H. Schuler, Radialion Research, 3, 58 (1955).
IC 1 -
c tc=c
, (18) C. E. McCauley and R. H. Schuler, J . A m . Chem. S O ~ .79, 4008 (1957). (19) N. Bach, "International Conference on the Peaceful Use of Atomic Energy," Vol. 7, United Nations, N. Y., 1956, p. 538.
20
SHEFFIELD GORDON, A. R. VANDYKENAND T. F. DOUMANI
I t was shown that approximately 60% of the hydrogen yield was independent of temperature and scavengers. The formation of this hydrogen was attributed to unspecified molecular processes. Onehalf of this hydrogen was accounted for by unsaturation and the remainder by dimer forming reactions which were independent of temperature. The remaining 40% of the hydrogen yield was attributed to free radical processes involving H-atom abstraction. The mass spectrum of n-hexane shows mainly C4, CI and C2positive ion peaks with only very small Csand c6 positive ion peaks. The chromatograms, which can be regarded as liquid fragmentation patterns, of irradiated hexane liquid show a large amount of product which must result from the formation of c6 species along with smaller amounts of C2,Csand C4 species. The marked differences between the mass spectral data and the chemical data can be attributed to the effect of state. Based on the observed product distributions as
Vol. 62
a function of temperature and scavengers the following primary processes are proposed
-*
+
G(-hexane)
+ + +
CeHir C6HI2 Hz 2CaH14-e 2CsHia Hz CaHir -e *CaHla ET. CeHu -e -Cj .Ck c6I14 CjHzj CkHaw
1.5
(I)
2.0 2.0 1.6 1. O
(11) (111)
(IV) (V) The H. atom formed in process I11 is assumed to react with n-hexane as H
+
2
+ CaHi, +caH13 + Ha
(VI)
Process I1 may actually occur as an ion-molecule reaction20as suggested in See. 3.3.2. Acknowledgments.-The author is much indebted t o J. s. Balwit for his generous assistance with the irradiations, to E. H. Winslow for the gas chromatography and to his colleagues for many stimulating discussions. (20) D. 0. Schisaler and D. P. Stevenson, J . Chem. Phya., 23, 1353 (1955); 24, 92G (1956).
IDENTIFICATION OF PRODUCTS IN THE RADIOLYSIS OF LIQUID BENZENE1 BY SHEFFIELD GORDON, A. R. VANDYKENAND T. F. DOUMANI~ Argonne National Laboratory, Lemont, Ill. Received June 84, 1967
Liquid benzene was irradiated by Coo0 gammas in a cell which continuously extracted the non-volatile fractions during the irradiation as well as in a static cell irradiated by the radiations from spent fuel rod assemblies. The non-condensable products were found to contain a Cl2 fraction containing biphenyl, phenylcyclohexadiene, phenylcyclohexene and bicyclic compounds in which neither ring is aromatic. A C18fraction containing hydrogenated terphenyls as well as a higher molecular weight fraction were also found. No evidence was found for cyclic compounds with non-cyclic substituents. The results are explained in terms of a fixed fraction of excited benzene reacting with benzene to form two of the major products as well as to form bicyclic radicals which further react with benzene and other radicals.
Introduction Extensive investigations have been conducted on the nature and yields of gaseous products from the radiolysis of benzene liquid and vapor. From these studies attempts have been made to formulate possible mechanisms for the primary processes consistent with the rather limited experimental results. One such study3 utilized mixtures of COD6 and C6H6, and from the yields of the various isotopic products concluded that a consistent mechanism could not be reached by assuming a free radical process exclusively, and that an ultimate molecule process had to be invoked. Very little work has been done with the condensable products although the total yields of these products is an order of magnitude greater than the gaseous products consisting of hydrogen and acetylene. This situation is not too suyprising. The condensable products obtained in the past have been complicated mixtures of high molecular weight hydrocarbons very difficult t o work with because of complexity and reactivity with oxy(1) Baaed on work performed under the auspices of the U.S. Atomic Energy Commission. (2) Union Oil Company. Brea, California; Resident Research Aasooiste, Araonno National Laboratory, 1956. (3) S. Gordon and M. Burton, Disc. Faraday Sac., l a , 88 (1952).
gen. However, it has been clear for some time and many investigators have pointed this out, that real progress in understanding the mechanism of the radiolysis of benzene could only be achieved by a more complete knowledge of the nature of these condensable products. There has been some qualitative and some quantitative description of this residue but nothing extensive enough to allow anything more than speculative formulation of a mechanism. Patrick and Burton4 have made the most extensive study of the so-called “polymer” to date. Using 1.5 MeV. electrons from a Van de Graff they found as others have reported, that the yellow residue when the irradiated benzene was distilled off under high vacuum was not a single substance. The viscosity and average molecular weight increased with total energy expended in the system and the number of double bonds per benzene ring decreased with total energy. Also they report that over the range of energy input which they studied, the yield or G value apparently was constant, averaging 0.75, in agreement with the value obtained by Manion and Rur60n.~ The lowest average molecular weight reported by these workers (4) W. M. Patrick and M. Burton, J . Am. Chcm. SOC.,76, 2026 (1954). (5) J. P. Manion and M. Burton, THISJOURNAL, 66, 560 (1932).
c
v