Dec., 1946
ADSORPTION O F
HYDROGEN AND NITROGEN ON
ably about O.Sl(2. The main difficulty is in estimating the concentration by titration with relatively sniall quantities of solutions. I wish to express m y gratitude to Det videnskapelige forskningsfond and to Universitetets jubileumsfond for grant3 enabling me t o carry out these investigations and toward h-orsk Hydro-Elektrisk Kvzlstofaktieselskab for placing heavy water at my disposal. I also wish to thank Docent Dr. Frivold and Professor Dr. Hassel for helpful suggestions and for constant interest throughout the progress of this research.
[ COh'IRII3UTIOK
FROM THE C I f E M I C A L
cHROMIUM
OXIDE G~;L
2535
Summary X vacuum distilling apparatus for deuterium oxide, and a conductivity cell are described. The construction of the cell allowed measurements to be carried out in an atmosphere of hydrogen. Conductance measurements at 25' are reported on solutions of NaOH in HzO,and of NaOD in D 2 0 . The limiting values of the ion conductances of OH- and of OD- in ordinary and in heavy water respectively have been computed. ULIVERSITETET I RECEIVED JVLY 19, 1946
O S L O , NORW.4P
EVCISEERISC DEPARTMENT, JOHXS HOPKINS UNIVERSITY]
Adsorption of Hydrogen and Nitrogen on: Chromium Oxide Gel 8
BY P. H. EMMETT'AND MARTINCINES'~
X number of years ago the late J. Howard2con- of x a t e r . The gel was then filtered, and dried at 150"; g. was taken as a sample. tributed one of the clearest indications that the ac- 2.33 The adsorption and surface area measurements were tivated adsorption of hydrogen on chromium- made in a standard adsorption apparatus that has already oxide gel was really a surface adsorption and not heen described.8 The dead space in the adsorption bulb an activated diffusion of hydrogen into the par- was calibrated in the usual way with pure helium at the temperature of each run. Dried tank hydrogen and preticles of gel to form a solid solution. He found purified tank nitrogen were used as the adsorbates. that the activated adsorption of 75 cc. of hydrogen on 87.5 g. of gel caused a decrease of about 50% Results and Discussion in the physical adsorption of hydrogen a t -78" The nitrogen adsorption isotherm for the sample deci-case in the physical adsorption of 0'. However, at the time his experi- of gel as measured a t -195" is shown in Fig. 1. ments were performed there was no good means The B.E.T.4 plot indicates an area of 310 sq. meavailable for judging the surface area arid the size ters per gram. The nitrogen monolayer on the of the pores in the gel. His results could only be 2.33-g. sample contained 165 cc. of nitrogen interpreted reasonably by assuming that much of (S.T.P.). The isotherm is the normal S-shaped the adsorbing surface was located in pores so small that they could bc coinpletely blocked by a layer of hylrogeri hcltl tightly b y activatrtl :itlsorption. a Since very few exaiiiples have come to light of 5 the iuhibiting 1:ffect of activated udsorption on rr; 250 physical :idsorption md, since ;L method is now -1 a~ailable~ for, ~measuring ,~ the surface area and judging something as to the pore size,6it seemed 200 worthwhi'le to repeat his results on :i similar saniple of chrom'ium oxide gel. 150 Experimental Procedure
-
A samplc of chromium oxide gel was prcparctl by the method used by Howard and originally d Lazier and Vaughen.' T o a liter of 0.04 nitrate was added $lightly more than a liter of 0.12 .\I ammonium hytlro:eIit .?.cldrt. - ' \ l e l l m In,titiile c,i I n < l u ~ t r i , + lKeiearcli, Pittsbnrgh. Pa. ( l a ) Presctit A i i d r e ~ . . Phillips Petroleum C v m p a n y , Burger, Texas. ( 2 ) Howarql, l'raiis. Filrodas .?,IC., SO, 278 (1934). ( 3 ) E m m e t t :and Bruiiauer, Tirxs J O l l R R A ~ . ,59, 1.753 (1937). (4) Brunaiirr, linirnrtt and 'leller, i b i d , 60, :109 (1938). ( 5 ) E m m e t t . Iiid. L?IC. Chem., 31, 639 (1943). (6) Iiie.;. ' J a n Nc~rdstr;ind. Johnson a n d Dauermeister, Tms J O I ~ R X U , , 67, I ? $ ? (l!l-t,~~), ( 7 ) Lnzier . i n t i Yauglicn. ibid., 54, 3080 (1932).
10U
0.1
0.2
0 8
0.4
0.i
0.i;
PIP,. Fig. 1.--Influence of chemisorbed hydrogen 011 the phpsical adsorption of S, a t - 195' on chromic oxide gel: 0, S?: adsorption before hydrogen chemisorption; 7, Sz adsorption at - 195" after chcrnisorptioii o f X.:l ('c. of hydrogen at 150 ". (8) Emmett, "Am. Sot. 'l'csting hlaterisls. Symposium un h-erv Methods fur Particle Size Detcrmiuatioti," 1!J-11, p. 113.
2536
P. H.
EMMETT AND M A R T I N CINES
variety characteristic of solids whose pores are much larger than molecular diameters. Since we wished to determine the influence of chemisorbed hydrogen on the physical adsorption of both hydrogen and nitrogen, adsorption measurements were made for these two gases a t various temperatures before introducing any chemisorbed hydrogen. The results are plotted in Fig. 2. The open circles represent, the adsorption for hydrogen a t -193', hydrogen a t -Bo, and nitrogen a t 0". The isotherms a t -78" for hydrogen and 0" for nitrogen are approximately linear with pressure.
Vol. 08
that his gel contained ,smaller pores and, therefore, might respond differently from ours, as regards the inhibitive effect of hydrogen. Perhaps his gel had smaller pores than our sample as a result of having been dried up to 400', whereas ours was dried to only about 150'. Consistent with this explanation is the fact that his hydrogen isotherm a t -78" and nitrogen isotherm at 0" were concave to the pressure axis whereas these isotherms on our samples are almost linear. This possible difference in pore structure seems to be the only reasonable explanation of the observed discrepancy; certainly there seems to be no question but that his experiments were carej 5.0 '('(' fully done and probably represent correct iso+ A therms on his sample of chromium oxide gel. G4.0 Howard2used his findings to prove that activated adsorption of hydrogen on chroniium oxide ON' gel was a surface adsorption rather than a solu3.0 tion effect. As he pointed out other adsorption S M and catalytic observations also seem to establish 2 2.0 the existence of activated adsorption as a surface 20 phenomenon. For example, the activated ad2s Xl.0. sorption of hydrogeng on iron catalysts strongly inhibits the activity of the catalysts for the inter>0 conversion of ortho-para hydrogenla a t liquid 0 100 200 300 400 500 600 700 nitrogen temperatures. This could hardlyeceur Pressure, mm. if the hydrogen held by activated adsorption were Fig. 2.-Influ.ence of chemisorbed hydrogen on the physical adsorption of hydrogen a t -195.8"; HP a t -78.5'; located elsewhere than on the surface of the catalyst. As a matter of fact, Howard's results, unless and NO a t 0". Circles represent adsorption before putting are perfectly 8.31 cc. of &ernisorbed HZon the gel; squares show the they can be shown to be in good evidence of pore blocking by hydrogen that adsorption after the chemisorption of 8.31 CC. of Hz. is held by activated adsorption and hence of the Solid circles and squares are desorption points. existence of this hydrogen as a surface adsorption To test the effect of chemisorbed hydrogen, S.31 on the particular ;ampie of chromium oxide that cc. was adsorbed a t about 130'. This was the net he employed. Our results merely indicate that for adsorption after quickly degassing the sample a t chromium oxide prepared and dried under the 130°, using the gas buret as a Topler pump. conditions used in the preparation of our sample, This adsorption is even larger, per gram, than hydrogen chemisorption does not inhibit the physithe one used by Howard. He chemisorbed 75 cc. cal adsorption of either hydrogen or nitrogen. on 37.3 g. of gel. Summary After the chemisorption of the hydrogen, physiThe chemisorption of 3.57 cc. of hydrogen per cal adsorption runs for hydrogen were made a t gram of chroniiuni oxide gel has been found to - 195 and -78'; runs for nitrogen were made a t have substantially no influence on the physical -195 and 0". The results after the hydrogen chemisorption are shown by squares in Figs. 1 and adsorption of hydrogen a t - 195 and -7S", and of 2 . It is apparent that the Chemisorbed hydrogen nitrogen a t 0'. The difference between these recaused no detectable decrease in the adsorption of sults and the marked inhibition o€ physical adeither hydrogen or nitrogen a t any of the tem- sorption reported by Howard is believed due t o differences in the pore sizes of the two gels. peratures or conditions employed. PA. RECEIVED Avc,usr 27, 1940 The cause of a discrepancy of this kind is dif- PITTSBURGH. ficult to establish with certainty. In spite of the (9) Emmett and Harkness, TIXIS JOURNAL, 57, 1831 (1935) similarity of procedure there is some indication (10) Emmett and Harkness, ibid., 6'7, 1628 (1935).
-
-~
a;
-q
-
