2068
NOTES
‘/bo that observed with alumina or silica; hence the analysis of the recorded data was more uncertain. No data of heats of immersion of alumina in water were found in the literature, for comparison with our results. For silica, the immediate heat was of the same order as that previously reportedsss for high-area silica. Immediate heat values for Ti02 and Graphon are in satisfactory agreement with published results.lo.11 Discussion.-The very small relative magnitude of the slow heat, e.g. as reported above for silica, explains why this effect has not been noticed before. A run ordinarily lasts about 20 minutes (Boyd and Harkins4 ran 16 minutes). For a first-order process, one must run two half-lives to get the reaction 75% to completion.la For our silica, with ha = 6.3 erg/cmS2and t l / , = 34 minutes, only 2.2 erg/cm.2 of slow heat would be evolved in a 20-minute run. This is 1.3% of the total heat, a magnitude that would ordinarily be laid t,o experimental error. When a twin calorimeter is not used, it is difficult to attain sufficient stability and low noise level to be confident of an effect of this size. A possible mechanism for slow evolution of heat for AlzOa and Si02 is rehydration of surface aluminoxane or siloxane groups to aluminol or silanol groups. Young13 has showed that silica dehydrates reversibly above 150” and (in part) irreversibly above 400°, but complete dehydration does not occur until above the sintering range (>goo”). Young discussed mechanisms but not energetics. Probably the hydrolysis of unstrained surface siloxanes should be endothermic. But a t sites of surface strain14 (e.g., highly strained’ siloxane groups) hydrolysis could be highly exothermic. Makrides and HackermanQ discuss hydrolysis of surface siloxane groups and comment on their heats of immersion of SiOz, “An undetermined, but probably small, heat evolution from rehydration is included in the apparent heat of wetting of powders evacuated at temperatures above 200-300°.” Thus it seems reasonable to attribute the slow heat for Also3 and Si02 to this hydrolysis. This mechanism agrees with the trend of increase of slow heat, and of h12, with activation temperature for alumina. The rather large probable errors in ha and t l l r are not surprising, since it might be hard to reproduce the exact concentration of strained groups, let alone the distribution of strain energies, in independent activations. Very probably the number of sites taking part in the processes leading to slow heat is much smaller than the number of sites responsible for the immediate heat. ( 8 ) F. E. Bartell and R. M. Suggitt, THISJOURNAL, 68, 36 (1954). (9) A. C. Makrides and N. Hackerman, ibid., 63, 594 (1959). (10) J. J. Chessick, F. H. Healey, A. C. Zettlernoyer and G. J. Young, ibid., 68, 887 (1954).
(11) P. Basford, C. Anderson, F. Murphy and G. Jura, “Proc. 2nd Int. Cong. Surf. Act.,” Vol. 2, Academic Press, Inc., New York, N. Y., 1957, p. 90. (12) J. M. Sturtevant, in Weissberger’s “Technique of Organic Chemistry,” Vol. 1. 2nd ed., Interscience Pub. Co., New York, N. Y.. 1949, pp. 731-845. (13) 0. J. Young, J . Coll. Sci., 13, 67 (1958). (14) E. B. Cornelius, T. H. Milliken, G. A. Mills and A. G. Oblad, THISJOURNAL, 69, 809 (1955).
Vol. 63
It is not clear why Ti02 produced effectively no slow heat. If the rehydration mechanism is correct for A1203and Si02, then the absence of slow heat for anatase and rutile may be due to the fact that they are formed as crystals from solution, and not by dehydration of a gelatinous hydrous oxide. Or perhaps if surface titanol groups exist, they can be dehydrated without leaving a strained group susceptible to hydrolysis. For graphite, the mechanism is probably quite different. Oxidation of surface carbons by water was suggested by McBain, et a1.,16 and was more recently studied by Pierce, et aL16 The reaction was slow, and the products were Hz, CO and no doubt chemisorbed oxygen. Presumably only carbons located at edges of basal planes would react with water. (Attack on carbon atoms in the interior of a plane would be endothermic, having a very large activation energy.) An estimate using bond energies, for possible reactions of edge carbons, gives a heat of 5 to 20 kcal./mole. Combining this range for the heat with the estimatelo that about 1/1500 of the sites for Graphon are hydrophilic, a value of about 1 to 4 erg/cm.2 is obtained. This is of the same order of magnitude as the observed values, which ranged from 5 to 13 erg/cm.2 Support of the U. 8. Air Force, under contracts AF 33(616)231 and AF 33(616)2824, is acknowledged. This work forms part of WADC reports 55-44 and 56-188. We thank L. A. Girifalco and J. A. McAndrews for the surface area measurements. (15) J. W. McBain, J. L. Porter and R. F. Sessions, J . A m . Chem. SOC.,6 6 , 2294 (1933). (16) C. Pierce, R. N. Smith, J. W. Wiley and H. Cordes, ibid., 78, 4451 (1951).
TRACER-DIFFUSION COEFFICIENTS OF CESIUM ION IN AQUEOUS ALKALI CHLORIDE SOLUTIONS AT $25’ BYREQINALD MILLSAND L. A. WOOLF~ Department of Radiochqmistrv, Research School of Phvsical Sciences, The Australian Nataonal Unzuerszty. CanbeTra, A.C.T. Received Juns 6 , 1060
Considerable data have now accumulated for the tracer-diffusion of several monovalent ions, in the three supporting electrolytes KC1, NaCl and LiC1, over the general concentration range 0.1-4 M . The ions, the coefficients of which have been measured under these conditions, are the anions C1and I- 8 and the cations Na+, 4--6 Rb+B and Hf.l All these measurements have been made by the reliable magnetically-stirred diaphragm cell method, which, however, sets a lower limit on the concentration, at which measurements are valid, of about 0.1 M . This limitation precludes the use of these (1) Department of Chemistry, University of Wisconsin, Madison, Wisconsin. (2) R. Mills, THISJOURNAL, 61, 1631 (1957). (3) R. H. Stokes, L. A. Woolf and R. Mills, ibid., 61, 1634 (1957). (4) R. Mille, ibid., 61, 1258 (1957). (5) R. Mille, J . A m . Cham. SOC.,71, 6116 (1955). (6) R. Mills, THIEJOURNAL, 63, 1873 (1959). (7) L. A. Woolf, Ph.D. Thesis, University of New England, Armidale, N.S.W., 1959.
c
I
Dec., 1959 1.0
1
0.9
NOTES
2069
Results The results of the work are given in Table I. The estimated accuracy in D is 0.4'%. Docs+has been calculated to be 2.057 X 10" cm.2/sec., taking hocs+= 77.26/ohm/cm.13 (13) Ref. 11, p. 452.
TABLE I COEFFICIENTS OF CS -k I N AQUEOUS ALKALICHLORIDE SOLUTIONS AT 25"
TRACER-DIFFUSION
0.8
DXlW cm.a/sed.
D/Do
0.1 0.5 1.0 2.0 3.0 4.0
KC1 1.981 1.957 1.951 1.897 1.810 1.707
0.963 .951 .948 .922 .880 .830
0.1 0.5 1.0 2.0 3.0 4.0
NaCl 1.936 1.892 1.828 1.651 1.504 1.315
0.941 .920 .889 .803 .731 ,639
0.0805 0.4025 0.805 1.610 3.221 3.990 4.025
LiCl 1.890 1.801 1.744 1 .607 1.351 1.216 1.208
m0f;sp. 0.
0.7
0.6
0.5
data for the testing of the Onsager limiting law.* Measurements in the very dilute region are, however, now being made with the continual monitoring capillary m e t h ~ d and ~ ~ 'data ~ in the concentration range up to 0.1 M for the above ions should soon be forthcoming. Such complementary data ought to allow formulation of theoretical and empirical extensions to the limiting law covering both dilute and concentrated solutions as in the analogous conductivity case.'l I n the meantime, accumulation of data such as are reported in this paper is justified for the intercomparison of cations and anions in the same supporting medium. In particular, the testing of semi-empirical viscosity corrections should prove very useful when the limiting laws can be extended. Experimental The potassium chloride and sodium chloride solutions were repared by weighing calculated amounts of the analytica? quality reagent and diluting to volume Lithium chloride was pre ared by a similar method to that of Stokes and Stokes.'* $he concentration of the initial lithium chloride stock solution was determined by conductance measurements after a propriate weight dilution and this then was diluted by vocme to give solutions covering the concentration range studied. These procedures were checked by conductance measurements with a Leeds and Northrup Jones bridge and shown to give an accuracy in concentration of better than 0.1%. Diaphragm cell manipulation has been adequately described in previous Papers.aI6 The radiotracer was Csla4 whch was obtained in the form of aqueous cesium chloride from the Radiochemical Centre, Amersham, England. As in previous studies,ap4 scintillation counting techniques were used. (8) L. Onsager, Ann. N . Y . Acad. Sci., 46, 241 (1945). (9) R. Mille and E. W. Godbole, A@&. J . Chem.. 11, 1 (1958). (10) R. Mills and E. W. Godbole, ibid., 12, 102 (1959). (11) R. A. Robinson and R. H. Stokea, "Electrolyte Solutions," Butterworth's Scientifio Publications, London, England, 1955, p. 153. (12) J. M. Stokes and R. H. Stokee, TRISJOURNAL, 60, 217 (1956).
0.919 ,876 .848 ,781 ,657 .591 .587
Discussion I n Fig. 1, the present data for Cs+ are compared with tracer-diffusion results for Rb+ taken from a previous publication.s The concentration dependence of the two ions is seen to be very similar and the values of the coefficients are in fact the same within the stated limits of error at all points of the curves. This might have been expected since both ions are probably unhydrated and of comparable size. This similarity means that DIDo (Cs+) can be compared to DIDo (Na+) in similar terms to DIDo (Rb*). The reader is therefore advised to consult the discussion of reference (6) for this comparison. Acknowledgments.-L.A.W. wishes to express his gratitude to the Australian National University for permission to use their facilities and also to the University of New England for leave to undertake the work. AN X-RAY DIFFRACTION INVESTIGATION OF SODIUM HYALURONATE' BY FREDERICK A. BETTELHEIM Chemistry Department, Adelphi Collew, Qarden Citu, N . Y. Received June 16, 1969
The purpose of this note is to elucidate the structure of crystalline sodium hyaluronate. Dif(1) Supported in part by a grant ((2-3984 BBC) of the National Cancer Institute. Public Health Servioe.
