Sept. 3, 1952
HEATC A P A C I T Y
O F METHANE
ADSORBED ON TITANIUM L)IOXIDE
4418
tion in the cycloheptatrienone and cyclopenta- exceed the observed much as the values calculated dienone rings is given by Brown's calculation l 2 for ionic structures tend to exceed such as may be of the number of a-electrons per atom by the observed,16 but show that the molecular moments method of Coulson and Longuet-Higgins. l3 From should be greater than those of ordinary aliphatic the charge distribution and the dimensions of the and aromatic ketones. The moment of the tetratropolone ring given by X-ray analysis,l 4 a dipole cyclone molecule is, presumably, increased above moment value 9.5 X 10-l8 may be calculated for that of cyclopentadienone by the resonance concycloheptatrienone. A similar calculation gives tributions from the twelve polar structures with for the hypothetical cyclopentadien- positive charge on a phenyl group. This would 5.6 X one, irrespective of whether the ring is taken as a account for the fact that the tetracyclone moment symmetrical pentagon or as resembling that of is not as far below the value calculated for the cyclopentadiene. The calculated moments greatly cyclopentadienone ring as is the cycloheptatrienone moment below its calculated value. (12) R. D . Brown, J. Cliem. S O L . ,2670 (1951). (13) C. A. Coulson and H. C. Longuet-Higgins, Proc. Roy. SOC. (London), Al91, 39 (1947) (14) J . hI. Robertson, J . L'heni. Soc., 1222 (1951).
lCOSrKIUUrION FROM
THE
(15)
Cf.C. P. S m y t h , THISJ O U R N A L , 63, 57 (lU41).
PRINCETON, KEW JERSEY
MORLEY CHEMICAL LABORATORY O F JYESTERN RESERVE Uh'IVERSIrY ]
Heat Capacity of Methane Adsorbed on Titanium Dioxide between 55 and 90'K. BY E. L. PACE,D. J. SASIVIOR AND E. L. HERIC RECEIVED NOVEMBER 1, 1951 Some features in the construction and operation of a low-temperature adiabatic calorimeter for the study of gas-solid systems are described briefly. The heat capacity of methane adsorbed on titanium dioxide (rutile) has been determined between 55 and 90'K. a t four coverages ranging from 0.2 to 1.0 monolayer. The heat capacity of the adsorbed methane is greater than that of solid methane and, a t 89'K., exceeds the value for liquid methane. The heat capacity-temperature curves point to the absence of any first-order phase transitions in the range of temperatures and concentrations studied. It is suggested that the relatively high heat capacity of the adsorbed methane may be due in part to a contribution to the heat capacity resulting from a transition from localized to mobile adsorption. Additional experimental data and theoretical considerations are necessary before a complete model for the adsorption can be proposed.
Many investigations have been directed toward the determination of the state of the adsorbate and the nature of phase transitions in the adsorbate for systems of gases adsorbed on a finely divided solid. The isotherm measurements of the n-heptane-silver, n-heptane-ferric oxide, n-heptane-graphite s y ~ t e m s l - by ~ Jura, Harkins and co-workers are particularly relevant to the present work. Analysis of the isotherms showed the presence of a firstorder two-dimensional condensation in the adsorbed film. The two-dimensional critical temperature was found to be approximately 0.5 of the three-dimensional critical temperature, a value in rough agreement with theory utilizing a simple model4for the adsorbed gas. It was a purpose of the present investigation to observe directly and evaluate quantitatively a two-dimensional condensation by using ;1low-temperature adiabatic calorimeter in measurements of the heat capacity of the adsorbed film in the 50 to 90'K. temperature range. Such a transition would appear as a discontinuity in the heat capacity-temperature curve. The choice of methane as a suitable adsorbate was based on the following considerations. Its three-dimensional critical temperature of 190'K. made i t reasonably probable that a two-dimensional condensation of the type observed by Jura and co-workers would occur in the temperature range to be studied. In addition, (1) G. Jura, E. H. Loeser, P. R. Basford and W. D. Harkins, J . Chem. P h y s . , 13,5 3 5 (1945). (2) G. Jura, E. H. Loeser, P. R. Basford and W. D . Harkins, i b i d . , 14, 117 (1946). (3) G. Jura, W.D . Harkins and E. H. Loeser. rbid., 14,344 (1946). (4) T.L. Hill, ibid.. 14, 441 (1946).
Hill4had pointed out that transitions from localized to mobile adsorption should occur a t low temperatures. The spherical symmetry of the methane molecule was a desirable feature if any interpretation of the results in terms of structure was to be made. Titanium dioxide in the rutile form was selected as an adsorbent after a preliminary study of the adsorption isotherms indicated that it was satisfactory from the standpoint of surface area, non-porosity, and reversibility of adsorption. The only comparable measurements are those of Simon and Swain5with the argon-charcoal system, Morrison and Szasz6 with the nitrogen-rutile system, and Morrison, et al.,7 with the argon-rutile system.
Experimental Apparatus.-The low temperature adiabatic calorimeter used in the experimental measurements followed essentially the same relative arrangement of cold reservoirs, radiatioii shields and calorimeter vessel as employed by Post, et uL.,~ in calorimetric studies of condensable gas systems. Because the activation of the adsorbent required the use of elevated temperatures with the calorimeter vessel containing the solid adsorbent, the construction of the calorimeter and the arrangement of the electrical leads, difference thermocouples, etc., was accomplished in such a manner as to allow convenient disassembly of the apparatus. Morrison and Lose have utilized a calorimeter of similar design in a study of the argon-rutile system. ( 5 ) F. Simon and R. C. Swain, Z . p h y s i k . Chcm., B88,189 (1935). (6) J. A. Morrison and G.J. Szasz, J. Chem. P h y s . , 16, 280 (1948). ( 7 ) 3. A. Morrison, J. hl. Los and I,. E. Drain, T r a n s . Faraday
SOL,4'7, 1023 (1951). (8) D. M. Y o s t , C. S. Garner, D . W. Osborne, T. R.Rubin and H. Russell, Jr., THISJOURNAL, 63, 3488 (1941). (9) J. A. Morrison and J. hZ. Los, Discussions Faraday Soc., Eo. 8, 321 (1950).
E I,.
4414
€'.ICE,
L).
J. SASMOR .IXD E.I,. HERIC
i n order to control accurately the temperature of the gas sample admitted t o the solid adsorbent, a portion of the monel iilling tube 60 cm. in length was cast in solder between two brass plates. The plates were supported by three S o . 25 B and S gage copper wires, 5 cm. in length, which were soldered both to the plates and the bottom of the liquid hydrogen container. i n the normal operation of the lorimeter, the cold drift along these wire supports was ompensated by the heat conducted from the room along the tube or generated by constantan heaters attached to the plates and the filling tube. Difference thermocouples, constructed of copper aiid constantan wire, Tvere appropriately tliffcreiices between the filling These temperature differences ing t h r admission of zhr gas sample or during actual thernial ~neasurenieiit~ iii order to t coiitleiis:ititrri nf tht. g:i< nutsitl(a t l i r caloririietri.
e calorimeter i-cascl \vas rnxiiit;ti~ied iii 311 adiabatic eiivironment hy nieans of ;i copper shield which totally enclosed i t . The construction arid operation ol this shield was similar t o that of Scott, et d.Lo The critical temperature differences existing between the top and side of the calorimeter vessel and the shield were indicated by fourjuxictioii, copper-constantan thermocouples connected to Leeds and Xorthrup Type K galvanometers. Temperature differences as small as 0,002' vere readily ohservable. The measurement of energy input t o the calorimeter vessel coiitainiiig thc gas -solid system duriiig a heat capacity tictermination and of the equilibriuni ?emper:ittire IWS ac'coniplisheti by appropri:tte electric:il circuits utilizing a ii-hite double potentiometer. Tlic time of the heater period t ~ a smeasured with :I synchronous interval tiiiirr operated with an accurately controlled 60-cycle source. 'The teuiperaturc of tht: system was read tvith a four-lead strain-free platinum resistance thermometer, constructed arid encased in a platinum thimble by the Leeds arid Sorthrup Company, and calibrated in the 1 1 t o 90°K. int fiureau of Stantiartis. i t iic.t~-poiiitr iiiately 25.5 ohnis. The quantity of gas \vas iiiehisured \sit11 the usu:il voltinictric and manometric apparatus. Pressures were tletermined with a Gaertner high-precision cathetometer. Materials.. --The adsorbent used in the investigation was a 48.j-g. portioii of a 20rJ-g. sample of high area rutile (TiOg) provided by the National Lead Company. The monolayer concentration as determined by the BET method was allproximately 18 ml. ( S T P ) of methane per gram. The adsorbate vas Research Grade methane of 99.5% purity obtained from the Phillips Petroleum Company. Thib niethanc was further purified by condensation in :i liquid air trap and rejection of the first and last fractious. Method.--The heat capacity of the titanium dioxide 1 rutile) anti of the methane-titanium dioxide system was cstablished ivitli a pressure of helium in the calorimeter vessel of [O.t'j nini. in order to hasten the distribution of energy. Thcriual eyuilibriurn \\-LIS thus obtained in about ten minlites after t h e input of energy. No significant adsorption OF hcliuiii hv the titaiiiuin dioxide nor measurable change ill t h e heat capacity of the titanium dioxide could he detected
thc tritc totel hcut valxicity, r l E *'
