PI

The National Research Centre, Cairo, Egypt. Received March ... In a mixture containing 83 mole % of water the relaxation time is about double that of ...
0 downloads 0 Views 437KB Size
40

A. R. TOURKY, H. A. RIZKAND Y. M. GIRGIS

Vol. 65

THE DIELECTRIC PROPERTIES OF WATER IN DIOXANE BY A. R. TOURKY, H. A. RIZKAND Y. M. GIRGIS The National Research Centre, Cairo, Egypt Received March PI, I860

The apparent solution moment of water in dioxane is 1.093 f 0.03 D. The electrical anisotropy of water in dioxane tenda to attain that of pure water at concentrations higher than 83 mole % of water. From polarization measurements, the coordination number in liquid water appears to be about 6, which would lead to a value of 1.86 D for the calculated vapor moment by Kirkwood's equation. In a mixture containing 83 mole % of water the relaxation time is about double that of pure water, and in 60 mole %, where there are six water molecules for every four dioxane molecules, the enthalpy of activation for viscous flow is nearly equal to that of water in the pure state.

Introduction Williams' determined the dipole moment of water in dioxane as well as in benzene. His values of 1.9 and 1.7 D, respectively, were considered nearly identical so that dioxane was assumed t o be just as good a solvent as usual non-polar solvents. The values 1.89, 1.90, 1.86 and 1.91 D also have been found by other investigators2-s as the dipole moment of water in dioxane. On the other hand, h e r l o f and Short,6 Wyman (quoted by Scatchard and Benedict'), and Hasted, et aZ.,* found that the static dielectric constant of water falls linearly and rapidly with the addition of dioxane. This phenomenon in which the molar decrement is about -7.7 a t 25", and also the lengthening of the relaxation time of water in the presence of dioxane were attributed by Hasted, et aL18to the formation of a hydration sheath bound by a 0-HO hydrogen bond. The object of this investigation is t o study, besides the determination of the dipole moment of water in dioxane, the anisotropic and the association factors, and to evaluate the relaxation time of water and the enthalpy of activation for viscous flow a t different concentrations in dioxane.

Experimental Pure dioxane and conductivity water were prepared according to recommended procedure~.~J~ The measurement of the dielectric constant, density, refractive index, viscosity, and the calculations were made as described previously.ll

Results and Discussion The electric moment of water in dioxane, Pd, calculated from the average or extrapolated value of P2m(Table I) is 1.93 =k 0.03 D. This value is higher than the value 1.7 D obtained by Williams' using benzene as inert solvent, and also higher than the true gas moment value, pol 1.84 D.l2JS This indicates that in dioxane there is an additional polarization resulting from the hydrogen bonding (1) J. W. Williams, J . A m . Chem. Soc., 82, 1838 (1930). (2) A. J. Weith, M. F. Hobbs and P. M. Grose, iW,, 70, 805 (1948). (3) E. P. Linton and 0. Maass. Can. J . Ree., 7 , 81 (1932). (4) P.Abadie and G. Champetier, C . r. acad. Set., 200, 1590 (1935). (5) Y. L. Wang, J . physik. Chem.. 45B, 323 (1940). (6) G . Akerlof and 0. A. Short, J . A m . Chem. 50%68, 1241 (1936). (7) G. Scatchard and M. A. Benedict, ibid., SS, 837 (1936). (8) J. B. Hasted, G. H. Haggis and P. Hutton, Trans. Faraday Sac., 4'7, 577 (1951). (9) W. Weissberger and R. Proskauer, "Organic Solvents and Methods of Purification," Oxford at the Clsrendon Press, 1935,P. 139. (IO) P. Thiessen and K. Hermann, Z . Elektrochem., 43,66 (1937). (11) A. R. Tourky, H. A. Rizk and Y. M. Girgis, THISJOURNAL, 64, 565 (1960). (12) R. Sanger, P h y s i k . Z., 81,306 (1930). (13) J. D.Stranathan, Phu.9. Rev., IS, 538 (1936).

TABLEI THEDIPOLE MOMENT OF WATERIN DIOXANE pia

a12

Wt

ciz

(g./om.l)

0 0.0087291 .022298 .033231 .056960 .071679

2.2005 2.4273 2.7798 3.0638 3.6803 4.0627

1.02229 1.02454 1.02548 1.02553 1.02682 1.02752

(om.')

ni-~

PZ (cm.8)

rs- D (cm.1)

Temp. 30'

Pam (Graph.)

Pam

--

E--- PI wa

4.40 79.268

z?l-D 7 = W1

1.4179 0.27957 1.4174 .31467 4.30057 0.15595 1.4163 .36311 4.04437 .I7559 1.4155 ,39742 3.82607 .I8566 1.4137 .45954 3.43927 .I8957 1.4127 .49160 3.23837 .19162

--

A

+

y(oa

prm - A + p i

A 4.13485 y pim (Palit-Banerjee) Prm P a m 9 79.399

0.18688, R

~ PD

-

-

p

-16.80404 4.41442 79.530

D 3.37,EPI = Ram 1.93 D.

0

3.27,

DPI = 3.60

Temp. 40" 0 2.1835 1.01120 1.4129 0.27977 0.0087291 2.3996 1.01373 1.4120 .31383 4.18147 0.16597 .022298 2.7353 1.01408 1.4110 .36140 3.94087 .I8696 .033231 3.0062 1.01463 1.4102 .39493 3.74507 ,18049 .056960 3.5935 1.01624 1.4098 .45627 3.37827 .I9545 .071679 3.9578 1.01673 1.4082 .48828 3.18877 .20063 A = 4.01658,y -15.79754 P2m (Graph.) 4.25. (PalibBanerjee) 9 4.29635 h m 76.467 Pxm 77.403 F s m = 76.985 TI-D 0.19429, RI-D = 3.50, EPI Rzm 3.40,DPI = 3.74 PD 1.93 D

-

9

- -

0

9

0

Temp. 50' 0 0.0087291 ,022298 .033231 .056960 .071679

2.1685 2.3714 2.6901 2.9469 3.5041 3.8497

1.00009 1.4080 0.27994 1.00163 1.4072 .31322 4.09244 0.14815 1.00218 1.4061 .35957 3.85134 .I8651 1.00282 1.4055 .39245 3.66554 ,18992 1.00351 1.4052 .45336 3.32464 .I9854 1.00492 1.4036 .48478 3.13764 .I9847 A 3.91661,y -15.09172 Pam (Graph.) 4.175, psm (Palit-Banerjee) 4.19655 P;m sm 75.218 Pam = 75.605 Pam 76.411 0.19347,RI-D = 3.48,EPa = Rim 3.38,DPI = 3.72 t3-D PD = 1.94 D

-

--

9

-

9

-

action and contributing t o the total polarization. The difference of about 0.1 D between the apparent solution moment of water in dioxane and the moment in the gaseous state is much smaller than the corresponding difference, 1.09 D,in the case of HCL2 This small increment in the case of water may be ascribed to its very low acid strength in comparison with that of hydrogen halides, so that its bond moment is only slightly affected by dioxane. The electrical anisotropy, 0, of water in dioxane as calculated from the Raman-Krishnan equation4 (14) C. V. Raman and K. 8. Krishnan, Proc. Roy. Sac. (London) 117~. 589 (192s).

Jan., 1961

THE DIELECTRIC PROPERTIES OF WATERIN DIOXANE

THE)ELECTRICAL ANISOTROPY, e,

AND THPJASSOCIATION RATIO,P p m / P p OF r

41

TABLE I1 DIOXANE h 0.041288 .lo034 .14391 .22802 .27410 .41869 ,59104 .70384 .80985 .83459 1 .00000

7

30’

0.32239 ,37237 .40757 .47186 .50519 .65166 ,78846 .84989 .I90643 .91609 .!36190

c-l/c+2----------. 40’

0.31815 .36647 .40071 .46368 .49645 ,65139 .78007 .84696 .go261 .91178 .96014

r

50°

30°

0.31373 .36036 .39356 .45494 .48716 .63532 .76926 .a4395

4.27 3.70 3.38 2.92 2.73 2.12 1.74 1.62 1.52 1.50 1.43

.89742

.go684 ,95830

(Table 11) first decreases rapidly with increase of water concentration up till 60 mole %, and then slowly tending to attain the value for pure water at a mole per cent. higher than 83. It may be noted that a t 60 mole % of water there are six water molecules for every four dioxane molecules. The vapor moment, yo, in the case of water as calculated by means of the unmodified Onsager equation16 using Wyman’s datal6 of the static dielectric constant of liquid is 3.08 D a t 30’, 3.07 D at 40°, and 3.05 D at 50’. Complex formation in liquid water would qualitatively account for the apparent increase in these calculated values of pol as compared with the measured vapor value 1.84,12.’3 or the dilute solution values 1.7-1.9 D.1-6 .4ccording to Kirkwood,” each molecule of water is surrounded by a shell of four nearest neighbors beyond which orientational effects do not extend and the H \O/H bond angle is 105*, so that the correlation parameter accounting for the hindered rotation amounts to 2.482. Hence, the calculated molecular dipole moment in water, y, from Kirkwood’s equation by the temperature-dependence method is equal to 2.66 D, and the calculated vapor moment, PO, is 2.12 D. The latter value evidently is nearer to the measured values in vapor or in dilute solutions than the values obtained by Onsager’s equation. As a rough measure of the association of water in dioxane at different mole fractions, Errera’s ratio, l7 Ppm/Ppl may be used. It will be seen from Table I1 that above 60 mole yo of water, the increase of this ratio with increase of mole fraction is greater than that a t concentrations below, and that the decrease of the ratio with increasing temperature is more pronounced at higher than at lower water concentration. illthough the Debye equation is of very dubious significance for a highly polar substance such as water, yet the reduction of the orientation polarization at unit mole fraction t u about one-sixth of its value at infinite dilution may be taken as an indiication that the hindered rotation in a spherical region surrounding a central molecule in liquid water is due to six neighboring molecules. It may be noted that in order to reconcile the longer (15) L. Onaager. J . Am. Chcm. Soc., 68, 1486 (1936). (16) J. Wyman, ibid , 58, 1482 (1936). (17) J. G. Kirkwood, J . Ckem. Phya., 7, 911 (1939); Ann. N. h a d . Sct.. 40, 315 (1940); Tian.3. Faraday Soc., M A , 7 (1946).

Y.

B

x

10:-

WATER AT DIFFERENT MOLEFRACTIONS IN -Ppm/Pp--

40’

60°

30°

40°

50’

4.29 3.73 3.42 2.95 2.76 2.10 1.75 1.61 1.52 1.50 1.43

4.68 4.08 3.73 3.23 3.01 2.30 1.91 1.74 1.63 1.61 1.53

1.03 1.10 1.16 1.30 1.38 1.48 1.91 2.42 3.08 3.29 5.48

1.03 1.09 1.15 1.28 1.36 1.41 1.85 2.30 2.98 3.19 5.35

1.03 1.09 1.14 1.27 1.35 1.41 1.82 2.24 2.90 3.09 5.19

distances of 0-0 and 0-H, which are about 3.1 and 2.85 A., respectively, in liquid water, as compared with the much shorter distances 2.76 and 1.01 A., respectively, in ice,lg with the higher assume that density of water, Panthaleon, et aZ.,20~21 water has a coordination number higher than 4, which was formerly assumed to be present in liquid water. The coordination number pointed out by these authors is 6 with four neighbors at 2.85 A. joined by hydrogen bonds and two other water molecules a t about 3.6 A. Accordingly, if this value 6 is taken instead of 4 as the number of neighbors which are coordinated around the central molecule in liquid water, a value of 3.223 for the correlation parameter, g! in Kirkwood’s equation will be obtained. This m turn will give a value of 2.33 D for the molecular dipole moment, p, in the liquid, and a value of 1.86 D for the gas or vapor moment, M. This latter value is in satisfactory agreement with the measured vapor value. Relaxation times calculated by means of Debye’s relation22using molecular radii and measured viscosities are in most cases not in agreement with those obtained from dielectric loss measurements. For this reason, this equation has been considered as an inadequate representation of the relation between relaxation time, molecular radius and macroscopic viscosity of the medium, and the inadequacy has been attributed either to the departure of the assumed mechanism of orientation from the actual mechanism, or the wrong substitution of the unknown internal friction coefficient by the experimentally measured viscosity. Consequently, in our solutions the relaxation time of water as calculated from this relation, using the measured viscosity coefficients, X (Table 111), and taking the radius of vater molecule1@ as 0.782 X lov8 cm., should be regarded as only approximate. Water in the pure state has a relaxation time calculated in this way equal to 0.807 X lo-” sec. at 30’, which seems to be in consistency with the measured value 0.83 X lo-’* (18) J. Errera. “Leipziger VortrHge,” 1929, p. 105. (19) J. A. A. Ketelaar, “Chemical Constitution,” Elsevier Publishing Co., Amaterdam, 1958, p. 420. (20) V. Panthaleon. € Mendel I. and W. Boog, Disc. Faraday Soe.,

81,200 (1957). (21) V. Panthaleon. H. Mend& and J. Fahrenfort, Natura, 181, 880 (1958). (22) P. Debye, Trans. Faraday Soc., 80, 679 (1934).

KANGYANG

42

see. at 25” by Collie, et U Z . , ~ and ~ Hasted, et aL8 In any of the mixtures investigated, the relaxation time of water evidently is higher than that in the pure state, and the lengthening in the time of relaxation increases with increase of water concentration till about 80 mole %, above which the relaxation time seems to decrease. In a mixture containing 83 mole % of water (corresponding t o 50 weight %) the relaxation time is nearly double that of pure water. The higher values of ‘T for water in the presence of dioxane relates, as previously suggested by other a u t h o r ~ ,to ~ ,the ~ ~ hydrogen bonding action, and may also be qualitatively accounted for by assuming that 7 is proportional to q/T. The temperature-dependence of the macroscopic viscosity a t each concentration up to 83 mole yo of water, is used to calculate the enthalpy of activation for viscous flow, AH,, from the slope of the straight line.24 From the data obtained (Table 111) it will be seen that in a mixture containing 60 mole (23) C. H. Collie, el al., Proc. Phya. Soc., 60, 145 (1948).

T’ol. 65

TABLE I11 THERELAXATION TI^, T, A N D ENTHALPY OF ACTIVATION FOR VISCOUS FLOW,AH,, OF WATERAT DIFFERENTMOLE FRACTIONS IN DIOXANE -7 I2

0.041288 ,10034 ,14391 ,22802 .27410 ,41869 ,59104 ,70384 ,80985 ,83459 1,00000

X l o 5 (poise)-. 40’ 50’ 30’ 1083 913 778 1104 919 776 1128 932 812 1179 975 816 1215 1055 828 1261 1069 888 1543 1269 1063 1666 1346 1123 1687 1353 1122 1601 1278 1038 801 656 549

-T

30’ 1.09 1.11 1.13 1.19 1.22 1.27 1.55 1.68 1.70 1.61 0.807

X 101‘ (set.)- A H , (kcal./ 40’ 50’ mole) 0.89 0.73 3.24 0.90 .73 3.33 -77 3.35 0.91 0.95 .77 3.37 1.03 .78 3.42 1.04 .84 3.43 1.24 1.01 3.70 1.31 1.06 3.82 1.32 1.013 3.81 1.24 0.98 4.25 3.71 0.638 0.516

yo of water AH, is nearly equal t80that in pure water, below this concentration AH, is lower, and above it AHv is higher than that of water in the pure state. (24) S. Glasstone, K. J. Laidler and H. Eyring, “Theory of Rate Processes,” McGraw-Hill Book Co., Inc., New York, N. Y.. 1941, chap. IX.

NITRIC OXIDE AS A RADICAL SCAVENGER IN THE RADIOLYSIS OF GASEOUS HYDROCARBONS BY KAKGYANG Radiation Laboratory, Continental Oil Company, Ponca City, Oklahoma Received March 85, 1960

Nitric oxide reduces energy yields of various products in the radiolysis of gaseous hydrocarbons. These yield reductions can be explained by assuming that nitric oxide scavenged thermalized radicals or that nitric oxide interfered with processes not involving thermalized radicals (such as ionic reactions). Attempts to determine what part of the nitric oxide effect must be attributed to the latter cause have not been successful so far. We have, however, found experimental evidence that as much as 10 mole % nitric oxide has a negligible effect on the product yields of the processes listed which do not involve thermalized radicals : (1)processes yielding hydrogen in the radiolysis of methane, ethylene, propylene and cyclopropane; ( 2 ) processes yielding ethylene in the radiolysis of propylene. These results show that, at least in certain situations, nitric oxide can be used to estimate unambiguously the thermalized radical contributions.

Introduction Scavengers are often used to estimate the contribution of thermalized radical processes in the study of radiolysis mechanisms. For unambiguous conclusions, one must be sure that these two conditions are satisfied : (1) The scavenger reacts with most of the thermalized radicals but has only negligible effects on the product yields of processes that do not involve radicah2 (2) Products from the scavenger-radical reactions have only negligible effects on the energy yields of radiolysis products. Previously, we have used nitric oxide as a radical scavenger in the radiolysis of gaseous hydrocarbons.’ Extensive work by Hinshelwood and his co-workers established that a sufficient concentration of nitric oxide (10 mole %) almost completely suppresses radical processes in homogeneous gas phase reaction^.^ The use of nitric (1) Pertinent literature is cited in K. Yang and P. J. Manno, J . Am. Chem. BOG.,81, 3507 (1959). (2) Here, and in what follows, the term “radicals,” unmodified, will refer to thermalized free radicals. The processes t h a t do not involve radicals include reactions of ionio and excited neutral species a6 well as hot atoms and hot radicals. (3) J. Jach and C . Hinshelwood, Proc. Roy. Soc. (London), A229, 143 (1955).

oxide to elucidate reaction mechanisms in gas phase radiolysis, however, requires careful consideration. The main difficulty is that a substantial concentration of nitric oxide (several per cent.) must be present a t the beginning of an experiment. This is necessary to permit carrying the radiolysis to sufficient conversion to obtain measurable amounts of products while still maintaining a high enough concentration of nitric oxide to scavenge most radicals throughout the experiment, Recent experimental results seem to indicate that, even with 10 mole % of nitric oxide, conditions 1 and 2 are approximately satisfied. For example, the contributions of radical processes estimated by nitric oxide inhibition data closely agree with the values estimated by isotopic methods where such comparisons are possible. Another example comes from a study of the ethylenetritium reaction initiated by P-decay of tritium.’ There it appeared that only nitric oxide-inhibited (4) L. M. Dorfman, THIEJOURNAL, 62, 29 (1958). (5) L. M. Dorfman, ibid., 6 0 , 826 (1956).

(6) G. G. Meisels, W. H. Hamill and R. R. Williams, Jr., ibid., 61, 1456 (1957). (7) K. Yang and P. L. Gant, J . Chem. Phys., 31, 1589 (1959).