Sept., 196%
IRRADI .ITIOP;
OF
ISOPROPYLBESZENE ADSORBED O N SILICA-ALUM IN^
mitchl8 potrtulate of a cage effect which favors recombination of larger radicals with a resultant de-
1625
gen yield ihan do methyl radicals to the methane yield, The observed general increase in liquid-
The smaller hydrogen atom may
radiolysis rields with increase in temperature and
escape the cage more readily, or hydrogen atoms may contribute much less to the gas-phase hydro-
partioularlg the much greater increase in G-{CHJ relative to G-(H,) is consistent with an increased
crease in yield.
(18) J Franck and E Rabinowtch, Tram Faraday Soc , 50, 120 (1934).
probability
Of
escape from the cage
Or Of
atom abstr,rction from the caging molecules.
IRRADIATION OF ISOPROPYLBENZESE ADSORBED ON MICROPOROUS SILICA-ALUMINA BY ROBERT R. HENTZ Socony Mobil
Oil Company, Inc., Research Pepartment, Pranceton, N . J . Recezved March 16, 1951
Irradiation with cobalt-60 ?-rays of a heterogeneous system consisting of isopropylbeneene and microporous silica-alumina has been studied over the entire range of isopropylbeneene electron fraction, F , and with two solids of different surface &rem. G(benzene) increases markedly relative to yields of all other products as F decreases and passes through a sharp maximum a t very low F ; yields for benzene formation azd isopropylbenzene decomposition are higher in the presence of solid than in the pure li uid a t the same temperature, 36 , even though the radiation is absorbed overwhelmingly in the solid. Characteristic G-v2ues for the major products a t low electron fraction are as follows for F = 0.0068: isopropylbcw e n e reacted, 3.8; benzene, 1.05; hydrogen 0.208; methane, 0.0042. Interpretation of the maximum as due to saturation of sites effective in isopropylbenzene dealkyiation permits a calculation for both solids of effective site concentrations which are in good agreement. The unique behavior of benzene yields and the higher per cent conversion of decomposed isopropylbeneene to benzene suggest that isopropylbenzene is chemisorbed on those sites effective in thermal dealkylation; that radiation produces in the solid electronic excitation of relatively long lifetime a t room temperature; and that this excitation energy is transferred to chemisorbed isopropylbenzene, where it is utilized ( a i t h greater efficiency than in the liquid) for the reaction of dealkylation to which the isopropylbenzene is predisposed on the chemisorption sites. A different mechanism is postulated for hydrogen formation reweighed after each high temperature evacuation, but was allowed to come to room temperature under high vacuum. The desired weight of isopropylbenzene then was introduced isopropylbenzene compares with catalytic and non- as in gas-phase radiolyses, and the cell was sealed off. At catalytic thermal cracking and whether a selective least 2 hr. was allowed for attainment of a homogeneous radiation dealkylation to benzene might be obtained distribution prior t o irradiation. Weights of solid ranged 0.2-40.0 g. and weights of isopropylbenzene from 0.01with reasonably high radiation-energy yield (G) in from 2.0 g. A new cell and fresh reagents were used in each the presence of a microporous solid which is a good experiment, except as noted. catalyst for the thermal dealkylation. A recent Irradiations were conducted in the same apparatus and paper1 on the first phase of this study reports re- in the same manner ag previously described.' After irradiation the cell was attached to the vacuum line, usually after sults on the cobalt-60 y-irradiation of pure iso- having stood overnight. The breakoff seal was broken and propylbenzene over a wide range of experimental gas products were removed from the solid in the reaction conditions. A communication to the editor has cell a t r w m temperature by repeated expansion through a been published2 on preliminary results obtained trap a t -118" (eth 1 bromide muah) into the 550 ml. volof a Saunders-?aylor* apparatus, followed by xqeawrein y-irradiation of isopropylbenzene adsorbed on ume ment in a 5.28 ml. volume, until no further gas could be microporous silica-alumina. removed. Liquid products then were removed from the solid by placing liquid nitrogen around the trap and a boiling Experimental water bath around the cell for about 4 hr. Any additional Chemicals.-The aame isopropylbenzene was used as in gas products (usually very little) then were removed before the earlier experiments,l and the same purification proce- isolating the cell by means of a stopcock and again after dures were followed. The microporous silica-alumina (10% stopaook closure by placing ethyl bromide mush around the alumina) beads used were from 2-6 mm. in diameter with a trap containing the recovered liquid, This liquid was transsurface area of 400 m.t/g., ore volume of 0.43 ml./g., and ferred by liquid nitrogen distillation into a weighed bulb on particle density of 1.15 g.&l. The low-surface silica-aluthe vacuum line, ,and the bulb was sealed off and reweighed mina was prepared by steam treatment (20 hr. a t 700") of the to obtain the weight of liquid recovered, When liquid rehigh-surface beads and had the properties: surface area, 169 covery was likely t o be very small, a weighed amount of m.2/g.; pore volume, 0.36 ml./g.; particle density, 1.25 g./ pure 'so ropylbenzene was added to the weighed bulb and ml degassetprior to transfer of the liquid products into the bulb. Procedures.-Prior to use, the silica-alumina was placed Gas products were analyzed mass spectrometrically in a muffle furnace a t 550" for 20-70 hr. and then stored in a and liquid products chromatographically. Except for isodesiccator containing anhydrous magnesium perchlorate. propylcyclohexane, products heavier than isopropylbenzene This procedure resulted in a weight loss of about 7'%. The were not found in sufficient concentration in the recovered desired amount of solid was weighed into a reaction cell of the liquid to warrant routine determination. This may be due type previously used in gas-phase irradiations'; the cell w a ~ partially to limitations of the chromatographic equipment attached to IZ vacuum line and placed in a furnace. The cell used but is more likely a result of inability to remove heavier then was maintained a t about 450" for approximately 22 hr. products from the solid by the technique necessitated. while evacuation proceeded to and was maintained at a presSince the silica-alumina solid used is a catalyst for thersure of lO-i-l0--6 mm. This treatment caused a further weight loss of less than 0.6%; therefore, the solid was not mal dealkylation of isopropylbenzene, it was found neoessary to employ a temperature for liquid product recovery -
This paper reports on the second phase of a study
undertaken to determine how radiation cracking of
.
(1) R.R. Hentz, J. Phys. Chem., 66, 1622 (1962). ( 2 ) R. R . Hentz, z b d , 6 6 , 1470 (1961).
(3) K. W. Saunders and H. A. Taylor, J. Chem. Phus., 9,610 (1941).
ROBERT R. HESTZ
1626
G (HYDROGEN) X 5
F
Fig. l.-y-Irradiation
of isopropylbenzene and high-siirface silica-alumina. .-
1.4
I
.z
I .o
,G (HYDROGEN) x 5 0.8 G. 0.6 LOW-SURFACE
RESULTS
0.4
0 2
-
I
LOW-SURFACE
A
RESULTS
50
F x to3
Fig. 2.-Comparison
of high-surface and low-surface results a t low F .
from the solid not much in excess of 100". Blank experimenta were performed in which benzene-isopropylbenzene mixtures and pure is0 ropylbenzene were adsorbed on the solid, allowed to stan: overnight a t room temperature (in one case after having been maintained a t 102" for almost 4 hr.), and then removed by the procedure described. Recovery of benzene proved to be quantitative and isopropylbenzene decomposition negligible in all cases; however, recovery of isopropylbenzene (and most likely, therefore, heavier products) is far from quantitative a t low weights of isopropylbenzene relative to weight of solid. Consequently, calculations of G(-isopropylbenzene) and per cent decomposition due to radiolysis are made by use of a rather large and not very reproducible correction for the amount of isopropylbenzene that remains adsorbed on the solid after the liquid recovery procedure. When weight of solid relative to that of isopropylbenzene is small enough for the correc. tion to be negligible, an upper limit for G(-isopropylbenzene) and per cent decomposition is obtained by ignoring the correction. In any case, these values have only an approximate significance. AE might be expected, the low-surface solid retains much less isopro ylbenzene and, therefore, yields more reliable values for &-isopropylbenzene). The fact that the low-surface values are consistently lower along with other observations suggests t h a t irradiated solid may retain more isopro ylbenz ene than the unirradiated solid which, if true, woulzmake the values for G(-isopropylbenzene) too large. Attempts to remove the liquid by saturation of the solid with pyridine followed by distillation from the solid at room temperature into a liquid nitrogen trap also gave quantitative benzene recovery but much poorer isopro ylbenzene recovery; therefore, this more difficult promlure was not used A blank experiment was peiformed in which isopropylbenzene and an amount of propene equivalent to the amount
T'ol. GG
of benzene that would be produced in an irradiation were adsorbed on and removed from the solid by the usual procedure. Only 0.3y0 of the propene w a ~recovered, and very small amounts of diisopropylbenzenes accounting for less than 8% of the added propene were found in the liquid recovered; however, little recovery from the solid of diisopropylbenzenes might be expected even if present in the adsorbed isopropylbenzene a t the theoretical 6% based on quantitative conversion of the propene. The significance of this result will be discussed later. Blank irradiations of the solid in the absence of isopropylbenzene yielded no gas product. Dosimetry.-Absorbed dose was determined with a ceric sulfate dosimeter as previously described,' with appropriate corrections for the electron density of the solid-isopropylbenzene system relative to 0.4 AI sulfuric acid. The results in Table I show that yields of all major products decrease with increasing absorbed dose. At the highest dose even the amounts (proportional t o G X dose) of methane, propane, and benzene have decreased. It was necessary to obtain accurately measurable G-values for comparable conditions over the whole range of composition studied in order to observe the dependence of G-values on F (electron fraction of isopropylbenzene in the syetem) without undue influence of the effect of dose dependence. This required, within the experimental limitations of this work, that a t the lower electron fractions the ratio of dose to weight of isopropylbenzene be kept approximately constant at a value near that of column 2, Table I. This value of the ratio, as shown by an average over many experiments, corresponds to approximately 20% decomposition of isopropylbenzene a t the lower values of F . At values of F greater than about 0.01 for high-surface solid the per cent decomposition decreased to a value of 47, for pure isopropylbenzene. Consequently, G-values for values of F greater than 0.01 may, if anything, be a little high relative to those a t lower values of F . Similar considerations apply to the low-surface solid. Nevertheless, Fig. 1 and 2 should represent quite well the important features of the dependence of G on F , especially in the low F region on whirh the interpretations are based. Obviously, as shown by the data of Table I, the G-values of Table I1 and of Fig. 1 and 2 are lower than the zero-dose values.
TABLE I DosE DEPENDENCE OF YIELDSIN IRRADIATION OF HETEROGENEOUS SYSTEM^
G*(Hydrogen) G(Methane) G(Propene) G(Propane1 G(1sobutane) G(-Isopropylbenzene) G(Bensene) G(Ethy1bensene) G(Isopropy1cyclohexane) 0 . 0 4 70 decomposition 16
0.003
0.03 20
-
e.v. X 10-21 10.5 51.8 175 0.150 0.093 0.061 ,0036 ,0024 ,0003 ,0007 ,0001 ,0001 ,0033 ,0013 ,0003 ,0025 ,0011 ,0007 .4 3.6 1.1 0.74 0.18 ,039 0.004 ,002 ,0007 .009 ,002 38 56 67
Dose, 1.24 5.31 0.194 0,208 ,0051 ,0042 ,0015 ,0004 ,0044 ,0033 ,0007 ,0011 14 3.8 2.3 1.05
High-surface solid a t 36' and electron fractions of isopropylbenzene near 0.007 (about 0.2 g. of liquid to 30 g. of solid). * Yield in molecules/100 e.v. based on total absorbed energy. In the experiment of column 4, Table 11, the solid was pumped overnight on the vacuum line after gas and liquid products had been removed. Then fresh isopropylbenzene was introduced and the system was irradiated. The results are shown in column 5 , Table 11. The reduced yields, particularly of benzene, show that the decrease in yields with increased dose is due largely to products which are not removed from the solid, even by overnight pumping.
Results and Discussion Introduction.-In addition to products shown in Table I much smaller amounts of the following products occasionally were observed in the presence of the solid: ethane, butane, butenes, isopentane, toluene, n-propylbenzene, and miscellaneous non-aromatics (C&,,). Acetylene and
IRRADIATION OF IBOPROPYLBENZEKE ADEORBED ON SILICA-ALUMINA
Sept., 1962
1627
energy requirements of the gas phase dealkybtion to beneene and propene for which ,IFo 0.57 e.v. and AHo = 1.0 e.v. at 25O. SILICA-ALUMINA The corresponding thermodynamic quantities are T a m p , ‘C not known for this reaction on the solid surface; 56 94 130 36 36C moreover, as will be discussed fully later, oi0 0218 0.00683 0.00761 0.00730 0 0244 F5 ,176 ,141 ,218 ,208 ,201 Gb(Hydrogen) chiometric reaction by which benzene is f in ,0127 ,0086 .0078 .0087 G(Methane) 0042 the radiolysis has not been established. se0033 0110 0186 0054 0050 G(Propane) quently, arguments from energetic considerations 0020 0071 0020 ,0006 G(1sobutane) 0011 that the solid must transfer energy to the adsorbate 65 2 1 28 1 66 G(I3enzene) 1 05 Yield in mole- are not conclusive; however, such results 8s the Electron fraction of isopropylbennene. cules/100 e.v. based on total absorbed energy. Irradia- foregoing strongly suggest that radiation energy tion of isopropylbenzene introduced to the solid of the ex- deposited in the solid is converted to a form which periment in the preceding column after liquid recovery. is efficiently transferred to the adsorbed isopropylbenzene, where it is used for benzene formation ethylene were not found. Whether a yellow poly- with an efficiency considerably greater meric residue was produced as in pure isopropyl- liquid phase radiolysis. Further, the pe benzene radiolyses’ could not be determined; how- of high yield, 1,13, to such a low electron ever, a major fraction of the product is left behind 0.00106, a t which 99.9% of the radiation energy is on the solid in some form. Columns 1, 2 , and 3 of initially absorbed in the solid and which correTable I1 show that with the possible exception of sponcjs to about 0.6% surface coverage (based 011 the hydrogen yield there is a general increase in a 50 A.2 cross section for isopropylbenzene) syggests product yields with increase in temperature. that transfer of energy from solid to ad Plots of major product yields vs. F (electron frac- rapid relative to the decay time of the re tion of isopropylbenzene in the system) for irradia- trander entity. The low-surf ace results tions a t 36’ are shown in Fig. 1 and 2 along w;th support these conclusions. corresponding “liquid lines” which represent the Complete surface coverage corresponds to about behavior expected if the absorbed radiation energy is initially partitioned between the two phases in F = 0.16 for the high-surface solid. At this elecproportion to their electron fractions (an explicit tron fraction one might expect a linear decrease assumption in all that follows) and if adsorbed iso- in G(benaene) with increasing P since the only propylbenaene behaves exactly like the liquid with- feature of the system that is changing is the ratio of unimolecularly adsorbed to total isopropylout enprgy exchange or interaction with the solid. Benzene Yield.-Figures 1 and 2 illustrate the benzene and, therefore, the partition of energy beunique behavior of the benzene yield. It is ap- tween these two different states of isopropylbend parent that the mechanism of benzene formation zene, each with a characteristic G(benzene) ; howis different from that for formation of other prod- ever, Fig. l and 2 show that the benze ucts. For high-surface solid the yield of methane reaches a maximum of about 1.45 a t F = does not greatly exGeed the “liquid line” value a t for the high-surface solid, then decreases rapidly to a low F and very s&xtly merges with the “liquid value of G(benzene) = 0.41 in the re ion F = 0.04line”; on the other hand, G(hydrogen) does not 0.17, and then decreases slowly to t e pure liquid vary greatly from the pure liquid value as F de- value. This passage through a sharp maxirpum at creases from unity to 0.00366-0.00683 when a sharp low-surface coverage suggests saturation a t the drop occurs; but, most significantly, the beiiaene maximum of those surface sites which are effective yield very markedly increases relative to the yields in benzene formation so that above F correspondof the other major products as F decreases. (Yields ing to the maximum additional isopropylhenzene for all products in irradiation of the pure liquid are occupies sites which compete for t,he transferred given in reference 1.) Indeed, benzcne becomes the energy but are relatively inefficient in beiiaene forgreatly predominant recovered product at low F mation. The maximum occurs at 1.0% surface whereas hydrogen predominates in radiolysis of coverage and corresponds to 2.0 X l O l a sites/m.2. the pure liquid, and the yields for benzene forma- Results for the low-surface solid tend to oonfirm tion and isopropylbenzene decomposition (G = this interpretation of the maximum. In this case, 1.8 for pure liquid I) are higher in the presence of the Fig. 2, a very sharp maximum occurs at about F = solid than in the pure liquid at the same tempera- 0.000616, corresponding to 0.8’% surface coverage ture even though the radiation is absorbed over- and 1.7 X 1OI8 sites/m.2. Agreement with the high-surface value is good. Oblad4 has presented whelmingly in the solid. data for a similar catalyst (12.5% alumina) from If 100% energy transfer from solid to adsorbate is assumed, then for the value of G(benzene) = which a value is obtained of 1.3 X ioi7 sites/m.2 1.13a t F = 0.00106 the yield of benzene from total for quinoline chemisorption. A mechanism is tentatively proposed as follows energy received by adsorbate also must be 1.13. If no energy transfer is assumed, then a yield of to account for the data on benzene formation. benzene based on only that radiation energy Isopropylbenzene is chemisorbed on those sites directly deposited in the adsorbate may be calcu- effective in thermal dealkylation. Radiation prolated as G / F = 1.06 X lo3, which gives a value of duces in the solid electronic excitation of relatively 0.09 e,v. (100F/G) of radiation energy directly long lifetime at room temperature. As suggested deposited in adsorbate for each molecule of benzene (*) G. -4.Mills. E. R . Baedeker, and A. G. Oblad, J . A m . Chem. Soc., formed. A comparison may be made with the l a , 1554 (1950). TABLE I1
-{-IRRADIATION
OF
IBOPROPYLBGNIENB AND
HIGH-SURFACEof isopropylbeneene E
a
i628
~ ~ O R I C KRT.
by Sutherland and Allen6 in discussion of their data
on radiolysis of pentane adsorbed on various solids,
these excited states may be electrons and positive holes which migrate independently through the solid. One of these entities may be trapped by isopropylbenzene on the surface and subsequently neutralized by the other with release t o the isopropylbensene of thc neutralization energy either as electronic excitation or as vibrational excitation of the ground d a t e (equivalent to a high local temperature as result of internal conversion of the neut,ralization energy into vibrational energy of neighboring atoms). Regardless of the detailed mechanism it is postulated that electronic excitation energy produced in the solid by irradiation is transferred to chemisorbed isopro ylbenzene, where it is partially ut,ilized to satisfy t e energy requirements of the dealkylation reaction to which isopropylbenzene is predisposed on the chemisorption sites. At higher temperatures (cf. Table 11)the average energy of adsorbed isopropylbenzene molecules is closer to the activation energy; consequently, a smaller fraction of the solid excitation energy is required for activation, and the probability of transfer or conversion of the requisite energy may be larger, It has been postulated that at an electron fraction correspofiding to saturation of chernisorption sites most effective in dealkylation, the benzene yield is a maximum since additional isopropylbenzene is adsorbed on sites at which excitation transfer occurs but with a lower probability of benzene formation. This may be due to a larger activation energy for dealkylation on these sites. The large decrease in yield of benzene with increasing dose may be attributed to a greater prgbabiljty for transfer of solid excitation energy to certain products which dissipate the energy without chemical effect or without benzene formation. I n addition to Caffrey and Allene and Butherland and Allen6 several other authors have invoked energy transfer from tl solid to an adsorbed phase to aceount for results on radiolysis of a heterogeneous system. Tw-o recent reviews are available.71s Clingman3 has reported a G = 2,2 for a more selective propane oxidation by X-ray irradiation in the presence of zinc oxide which absorbs essentially all the energy. He suggest8 that energy transfer by interaction of a positive hole with anionically adsorbed oxygen produces a new species of adsorbed oxygen that is a reactive ietermediate in the altered readion. Hydrogen Yield.- Whereas the benzene yield on high-surface solid decreases sharply at F less than 0.00177, the hydrogen yield begins to fall sharply in the region F = 0.00365-0.00683. It is possible that hydrogen formatioii is a higher energy process that occurs via shorter lifetime, higher excited states of the solid; however, the greatly reduced hydrogen yields on a solid whose surface area has
K
( 5 ) J. W. Sutherland and A. 0. Allen, J . A m . Chem. Soc., 8 8 , 1040 (19131). ( 6 ) J. M. Caffrey, Jr., and A. 0. Allen, J , f h y s . Chem., 6!2, 33 (1958). (7) M. Hamsinsky, Jadmna eneigze, 7 , 73 (1961). (8) R. Coekelbeigs, A. Crucq, and -4.Fiennet, Adean. Cataluszs, 13, 55 (1961). 19) W. H. Clingman, Jr., Ind. Eng. Chem., 6 2 , 915 (1980).
HESTZ
Vol. 66
been reduced by only a factor of 2 4 is difficult to explain in such terms (e$ Fig. 2). Sutherland and Allen5 have observed that replacement of 19% of the sodium ions of Linde Co. “Molecular Sieve 13X” with hydrogen ions resulted in a great increase of the hydrogen yield on irradiation of adsorbed pentane. These authors postulate that the hydrogen ions on the surface may trap electrons from the conduction band to form hydrogen atoms, which then may abstract hydrogen from pentme, 4 similar mechanism could account satisfactorily for the data 011 hydrogen yields from adsorbed isopropylbenzeiie. Without specification of the detailed mechanism for the bond rupture process, it is postulated that irradiation of the solid dissociates surface 0-H bonds and that the hydrogen atoms migrate on the surface until they are recaptured by a surface valence, add to the ring of an isopropylberiaene molecule, or form hydrogen by abstraction from the /3 C-H bond of an isopropylbenzene molecule. The latter two processes would have fixed relative probabilities as composition is varied; however, at a sufficiently low electron fraction of isopropylbenzene, recapture of hydrogen atoms by the surface would become important relative to reaction with isopropylhenzene, and the hydrogen yield would decrease with decreasing F . Above the eIectron fraction at which hydrogen yields decrease sharply, from about F = 0.003851.0, the approximate constancy of G(hydrogen) can be explained if it is assumed that G-values for hydrogen atom production from solid and from isopropylbenzene do not differ appreciably, Assuming one hydrogen atom from the solid per molecule of hydrogen produced by irradiation, the maximum number of hydrogen atoms obtained from a gram of solid in an experiment was 7 x lo1*. If the solid is assumed to retain only 1% by weight of water after the pretreatment procedures, there would be 7 X 102O hydrogen atoms present in a gram of solid. Thus, only 1% of the solid’s hydrogen atoms would have been rcrnoved in this experiment. In terms of the proposed mechanism for hydrogen formation, the much reduced yields of hydrogen from isopropylbenzene on low-surface solid are explained as due to destruction of labjle 0-H bonds by the surface reduction treatment.1° The failwe to obtain any hydrogen on irradiation of solid alone is consistent with this mechanism. Other Products.-Caffrey and Allen6 as well as Butherland and Alle1~have commented on the suppression of unsaturates in irradiation of pentane adsorbed on silica gel and the enhancement of branched products. In this work also nonaromatic unsaturates found in irradiation of the gas and liquid were either absent or considerably suppressed in the products recovered from the solid. Isobutaiie formation seems to be enhanced by the presence of the solid, and isopropylcyclohexane was found only in the presence of solid. Selectivity.-The percentage of decomposed isopropylbenzene which has been converted to benzene (per cent coiiversion) may be calculated from values for G(-isopropylbenzene) and G(benzene); however, as discussed in the section on (10) J. L. Weill, Chzrn. mod., 6, Xo. 37, 193 (1980).
Sept. 19621
ISOPIESTIC ISVESTIGATIOS
O F DI-(2-ETHYLWEXYL)-PHOSPHORIC4 C I D IS ?%-OCTANE
experimental procedures, values for G(-isopropylbenzene) are not very reliable because of the large correction necessitated by incomplete recovery from the solid. I n ten experiments with highsurface solid per cent conversion values ranged from 1 3 4 7 % . Seven of the ten values lay between 21 and 41y0, and the average of all ten was 29%. On low-surface solid the recovery of isopropylbenzene was much better, and no correction was used for isopropylbenzene retained; consequently, the values of G(-isopropylbenxene) used are too large and per cent conversions are minima. The range’ of per cent conversion from five experiments with lowsurface solid was 29-36%, with ai1 average of 37%. AppIication of a quite small correction (10-fold smaller than that measured for high-surface solid) for isopropylbenzene retention on the low-surface solid brings the average per cent conversion to near 50% and reduces the spread in values. It would appear that the per cent conversion on low-surface soljd is appreciably higher than that for highsurface solid. The much lower values for G(hydrogen) add support to this view; however, it is possible that retention by irradiated high-surface solid is greater than the retention correction used and, therefore, that per cent conversions calculated for high-surface solid are low. Conversion in liquid radiolysis is about 3%, and the maximum value in gas radiolysis would appear to be about 18%.1 There appears to be little doubt that presence of the solid has enhanced the selectivity of the radiolysis of isopropylbenzene considerably, as shown particularly by the results for per cent conversion with low-surface solid. Low vdues for G(propene), about 0.2% of G(benzene), may be due to failure to produce pro-
1629
pene or the inability to recover propene from the solid containing adsorbed isopropylbenzene, a8 discussed in the section on experimental procedures. Since thermodynamic equilibrium at loo”, the temperature used for recovery of liquid products, overwhelmingly favors isopropylbenzene relative to benzene and propene, observation of appreciable benzene yields is contingent on failure to produce equivalent amounts of propene or on conversion of propene to other products on the solid. Possible fates of propene are polymerization and formation of diisopropylbenzene. The possibility that propeiie is formed in an amount equivalent to benzene and subsequently disappears in isopropylbenzene alkylation would decrease the amount of G(-isopropylbenzene) attributable to side reactions and impose an upper limit on conversion of 50%; this value appears to be closely approached with low-surface solid. In the blank experiment (cf, Experimental Procedures) less than 8% of unrecovered propene was recovered as diisopropylbenzenes; however, this represents positive evidence for such a reaction, Thus, a radiation-induced, solid-catalyzed unimolecular split into benzene and propene with conversion of propene to diisopropylbenzene could be the major reaction occurrent. The unique behavior of the benzene yields and the higher per cent conversion of isopropylbenzene to benzene on a solid which is a catalyst for thermal dealkylation suggeet the possibility that a solid to some extent may direct absorbed radiation energy into that reaction for which it is a thermal catalyst. It is intended to test this speculation by the use of solids of varying catalytic activity, particularly a very pure silica of little or no catalytic activity,
AX ISOPIESTIC INVESTIGATlON OF DI-(2-ETHYLHEXYL)-PHOSPHORIC ACID (DPA) AND TRI-R-OCTYLPHOSPHINE OXIDE (TPO) I N n-OCTANE’ BY C. F. BAES,JR. Contribution from the Oak Ridge National Laboratory, Oak Ridge, Tennessee Received Februaw 88, 1968
Isopiestic measurements are presented in which octane solutions of di-( 2-ethylhexyl)-phosphoric acid (DPA) and tri-noctylphosphine oxide JTPO) are compared with triphenylmethane (TPM)-octane as the reference solution. In the range 0-0.2 m the results show deviations up to 10 and 20% from ideal solution behavior of DPA dimer and TPO monomer, res ectively. While partial trimerieation of DPA (in qualitative agreement with previous iron( 111) extraction results for 8 P A ) and partial dimerization of TPO (which is not supported by molar polarization results) can account in part for the resuIts, it is evident that non-specific non-ideal behavior of the solutes also is involved. Practical activity coefficients were , , !og estimated on two separate assumptions: ( 1 ) YTPI zz 1, giving log -I(DPA)* = - 0 . 5 2 2 7 m ( ~ p ~ , ~ ~ / 0a . 4 2 0 m ( ~ p ~and YTPO = - 1.168mTpo 0.149m~po2; and ( 2 ) log Y(DPA)% --0.6432m(~~a,,’/8(from iron(II1) extraction results), giving YTPM = -0 737m~pnrand log YTPO = -1.886m~p0 f 0 . 2 4 5 ~ ~ ~ ~ 0 ~ .
+
The isopiestic measurements reported here were undertaken in connection with solvent extraction studies of di-(Bethylhexy1)-phosphoric acid (DPA) 2-4 and tri-n-octylphosphine oxide (TPO). 5 , 6 (1) This communication is based on work performed for the U. S. Atomic Energy Commission a t Oak Ridge National Laboratory, Oak Ridge, Tennessee, operated by Union Carbide Corporation. (2) (a) C. F.Baes, Jr., R . A. Zingaro, a n d C. F. Coleman. J. Phys. Chem., 62, 129 (1958); (b) C. F. Baes, Jr., a n d H. T. Baker, ibid., 64, 89 (1960). (3) C. A. Blake, K. B. Brown, a n d C. F. Coleman, “ T h e Extraction’ and Recovery of Uranium (and Vanadium) from Acid Liquors with Di-
+
(2-ethylhexyl) -phosphoric Acid and Some Other Organophosphorus Acids,” ORNL-1903, M a y 1 3 , 1965, p. 106. (4) C. 8 . Blake, D. J. Crouse, C. F. Coleman, K. B. Brown, a n d A. D. Kelmers, “Progress Report: Further Studies of the Dialkylphosphoric Acid Extraction (Dapex) Process for Uranium,” ORNL-2172, Sept. 6, 1956, p. 110. ( 5 ) H. T. Baker a n d C. F. Baes, Jr., “An Infrared and Isopiestic Investigation of the Interaction Between Di-(2-ethylhexy1)-phosphoric Acid a n d Tri-(n-octy1)-phosphine Oxide in Octane.” paper presented a t ACS Meeting, Chicago, Sept. 7-12, 1958, paper in preparation. (6) C. A. Blake, I
