1586
Vol. 60
NOTES TABLE I
x 10'8, e.8.u.
Compound
a
P
El
VI (ml./g.)
Mr
PPO
Po
Cyclohexyl azide Cyclopentyl azide 3-Bromocyclohexene
4.9550 4.9827
-0.10449 -0.10487
2.2735 2.2769
1.1462 1.1464
35.37 31.18
155.65 138.69
118.51 105.95
2.37 i 0.03 2.27 f 0.03
4.1901
-0.32672
2.2741
1.1461
35.92
166.29
128.57
2.51 i 0.04
Cyclohexyl and cyclopentyl azides were prepared by the method reported by Boyer,3 from the corresponding bromides by refluxing with an alcohol-water solution of sodium azide. After extraction with ether and removal of the latter by distillation, the azide was distilled in an atmosphere of nitrogen under reduced pressure. Cyclohexyl azide distilled at 84-65' a t 21 mm.; d 2 5 0.98546; n Z 61.4690. ~ Cyclopentyl azide distilled a t 51-52' at 20 mm.; d% 0.9789; n z 51.4615. ~ An indication of purity was obtained by the constancy of refractive indices of successive distillation fractions. Cvclohexene was brominated bv refluxing with N-bromosucc'lnimide in carbon tetrachlor~de4to giye 3-bromocyclohexene. After removal of the succinimide by filtration, the carbon tetrachloride and excess cyclohexene were distilled off. The 3-bromocyclohexene was distilled under reduced pressure; b.p. 56-57' at 11 mm.,6 d2sa 1.3767; 12.261) 1.5265. The 3-bromocyclohexene decomposed so readily that it was necessary to redistil the compound, make the solutions and measure their dielectric constants and densities immediately and in rapid succession. Density.-The densities of the benzene solutions were measured a t 25.00 f 0.02' with calibrated 2 5 4 . Weld type specific gravity bottles. Smaller pycnometers were used for the pure substances. Dielectric Constants.-The dielectric constants of the benzene solutions of the compounds were measured a t 25.00 f 0.05' with the heterodyne beat apparatus and the cell described previously.6
p
bromide,1° 2.12 D, and t-butyl bromide,'O 2.21 D, show an increase in the moment as the branching in the compound increases within the sphere of induction. This work indicates that the dipole moment of a cyclic compound is closer t o that of a branched chain than to a straight chain compound. Since cyclohexene itself has a dipole moment of 0.75 D,ll the moment of 3-bromocyclohexene would be expected t o be greater than that of either cyclohexyl azide or bromide. (11) M. Puchalik, Acta Phys. Polon., 4 , 145 (1935).
THERMAL EFFECTS I N THE CHEMISORPTION OF OXYGEN ON NICKEL: A COMPARATIVE INVESTIGATION USING POWDERS AND EVAPORATED FILMS BY R. M. DELL,^^ D. F. KLEMPERER~~ AND F. S. STONE Department of Physical and Inorganic Chemistry, Uniuersify of Bristol, Bristol, England Received June 8, 1056.
A comparative study has been made of the behavior of nickel powder and nickel film toward oxygen chemisorption. It becomes evident that, besides cleanliness, other less commonly emphasized factors contribute markedly t o the observed differences in behavior. Nickel powder, prepared from the oxalate, was 3CUVlMl(EI 2)' reduced with hydrogen and outgassed.2 When in which E1 and V1 represent the dielectric constant Specimen I (wt. 5.66 g., S.A. 9.65 m.2) was exposed and specific volume of the pure solvent while M I to 1.5 mm. pressure of oxygen at -183", very and Mp are the molecular weights of solvent and rapid adsorption of 4.54 ~111.~ (S.T.P.) occurred. solute, respectively. The constants LY and p, E1 On warming to 20°, a further 1.07 ~ m was . ~adand 81in the empirical equations were evaluated by sorbed, Assuming equal areas of the three main the application of the method of least squares t o planes to be present, a 1 : l atomic monolayer of ~. the experimental data. The orientation polariza- oxygen on nickel requires 0.284 ~ m . ~ / m . Adopttion, PO,was obtained in the usual manner, using a ing this as a standard for comparisons, Specimen I correction of 5% for atomic polarization. The dipole adsorbed 1.66 atoms of 0 per Ni a t - 183", rising to moment was determined by the use of the equation 2.05 a t 20". Studies with other specimens, however, showed p = 0.01273(PoT)'/2 X 10-18 e.8.u. that the limits of the uptake are strongly dependent The constants, other data and the dipole moments upon factors other than temperature. Pressure calculated are listed in Table I. and porosity, for example, both have an influence A comparison of the dipole moments of cyclo- upon the uptake limit. The controlling factor is hexyl bromide,a 2.3 D, and cyclopentyl b r ~ m i d e , ~evidently the dissipation of the heat of adsorption. 2.20 D, with the analogous azides shows that the At the moment' of contact with oxygen, there is an bromo and azido moments are nearly identical. intense local disturbance in the surface layers, and However, n-butyl bromide,1° 1.85 D, sec-butyl unless the heat can be distributed rapidly, incor(3) J. H. Boyer, University of Michigan Engineering Research poration of adsorbed oxygen as oxide will be encourInstitute Reports, October, 1951. aged. New sites are then generated and in this (4) K. Ziegler, A. Spiith, E. Schaef, W. Schumann and E. Winkelway a number of layers of oxygen can be taken u p mann, Ann., 651, 80 (1942). by the solid. Under sufficiently extreme condi(5) A. T. Blomquist and J. Kwiatek, J. A m . Chem. SOC.,79, 2098 (19511, reported b.p. 68-70.5' a t 26 mm. tions, the effect appears as the well-known pyro(6) H. 0. Spauschua, A. P. Mills, J. M. Scott and C. A. MacKenzie, phoric property-Specimen I, for instance, was pyibid., 7 2 , 1377 (1950).
Results and Calculations The total polarization at infinite dilution, Pi, of each compound was calculated by the method of Halverstadt and Kumler' using their equation PZ' = (El - 1HEl 2 W z V 1 MlB)
+
+
+
+
(7) I. F. Halverrtadt and W. D. Kumler, ibid., 64, 2988 (1942). (8) J. W. Williams, ibid., 32, 1831 (1930). (9) M. T. Rogers and J. D. Roberts, ibid., 6 8 , 843 (1946). (10) A. Parts, 2. pRgsik. Ckem., B7, 327 (1930).
(1) (a) Admiralty Research Laboratory, Teddington, England; (b) Division of Tribophysice, C.S.I.R.O., Melbourne, Australia. (2) R. M. Dell and F. 8. Stone, Trans. Faraday SOC.,SO, 501 (1954).
Nov., 1956
NOTES
1587
rophoric when exposed t o 50 mm. pressure of oxygen a t 20O-but the significance of inadequate heat dissipation has seldom been considered in orthodox adsorption studies. Evaporated films may also be 0 , I expected to show the effect (see below), especially I1 when a thick film is deposited on a thin glass base. The high uptake limits for oxygen chemisorption \, I on nickel and iron films reported by Beeck3 are \ i probably to be attributed to this cause. h When coupled with poor heat dissipation, the + J ' Limitor high heat of oxygen adsorption (-100 kcal./mole) makes for difficulties in adsorption calorimetry. A j i direct determination of the heat of adsorption of oxygen on nickel powder is liable to a substantial error for this reason. The tendency is for hot spots to develop in the region of the specimen first encountered by the gas. Some improvement may be obtained by using oxygen-helium m i x t ~ r e s . ~I n the present case, however, the problem has been overcome by using nitrous oxide. NzO decomposes very slowly on nickel a t 20" yielding chemisorbed oxygen, but NnOitself and Nz are negligibly chemisorbed at low pressures. Using this r n e t h ~ d , ~ Specimen I1 (11.4 g., 43.5 m.2) gave the results , ~ whole shown in Fig. 1. (In contrast to c ~ p p e rthe of the nickel surface was active toward NzO decomposition.) Less accurate results, obtained using O z H e mixtures, are also included. Each heat determination with NzO increased the coverage by only 0.2%; the coverage range between determinations was traversed by admitting charges of oxygen. Bearing in mind the high heat and the irreversibility of the chemisorption, the constancy of the heat is attributable to a "layering effect." I I I Due to the slow speed of reaction, the small 0.1 0.2 0.3 0.4 0.5 amount of oxygen produced by N20 decomposition Vol. of oxygen adsorbed ( ~ m . S.T.P.). ~, during a heat measurement is probably chemisorbed Fig. 2.-Heat of adsorption of oxygen on evaporated nickel randomly on the adsorbent. On the other hand, (wt. 0.050 g., S.A.1.67 m.a) at 20°, using a Beeck-type the much larger charges of oxygen admitted be- film calorimeter.8 I n the absence of a conventional surface area tween heat measurements are likely to be adsorbed measurement, the area in this case has been assessed on the correpreferentially on the first portion of bare surface assumption that the adsorption limit a t 0.474 which they e n ~ o u n t e r . ~Each fresh increment of sponds to a monolayer (see text). gas (whether oxygen or nitrous oxide) is therefore gave a rapid adsorption of 0.179 cm.a, correspondadsorbed upon a sparsely covered surface, with the t o 1.17 layers. These values, which approxiresult that the observed heat of chemisorption re- ing t o a monolayer, are substantially smaller than mains constant. As complete saturation is ap- mate those observed with the powdered nickel specimens proached, the conditions change abruptly t o the (e.g., 2.05 and 2.96), a result which we attribute to other extreme: the heat then falls steeply to a value the much easier dissipation of heat in the films and characteristic of the heat of adsorption of oxygen on support. The heat of adsorption, however, is nickel oxidem2 For Specimen 11, saturation at 20" their not without effect. With Specimen IV corresponded to 2.96 layers. The use of N 2 0 apparently surface area was measured both before and after throughout would presumably have given a lower the chemisorption of the oxygen. It was found that uptake limit, but the very slow rate of decomposi- the the krypton v m decreased by 30% following the tion rendered this experiment impracticable. chemisorption of the oxygen layer. Our tentative As the work on evaporated films will be described interpretation is that the local mobility conferred in detail elsewhere,6 only certain salient features upon the nickel by the high heat of adsorption has will be reported here. Specimen I11 (0.020 g., caused sintering. 0.632 m.2), an evaporated film, was exposed to oxyT o complete the comparison, the heat of adsorp. ~ tion on films has been measured. I n view of the gen at 20" and 0.5 mm. pressure. 0.187 ~ m was adsorbed immediately, corresponding (4.v.) to better heat dissipation, direct measurement with 1.04 layers. Specimen IV (0.017 g., 0.540 m.z) oxygen could be made in this case. Some results are shown in Fig. 2. Fifteen admissions, each of (3) 0. Beeck, "Advances in Catalysis," Vol. 11, Academic Press, New York, N. Y., 1950. 0.036 cm.3, were made and residual oxygen pressures (4) W. E. Garner and F. J. Veal, J . Chem. Soo., 1436 (1935). only appeared in the last two increments. The con(5) R. M. Dell, F. S. Stone and P. F. Tiley, Trana. Faraday Sac., stancy of the heat as a function of coverage is again 49, 195 (1953). attributable to a layering effect ( g . ~ . ) . Compared I ( I D. F. Klemperer and F. S. Stone, to be published.
".
',
~
uptokc
NOTES
1588
with the powder, the high heats in Fig. 2 are to be expected. Aside from considerations of cleanliness, the condensed film is a very imperfect porous material with a high degree of heterogeneity. These heats must also contain a contribution from strain energy, for if the chemisorption confers upon the surface the ability to sinter, it must also relieve strain. Better values of heat of adsorption (and catalytic parameters) can obviously be obtained by using annealed films, but, as is well known, their small areas introduce new limitations. It is in these respects that parallel work with powders assumes its proper perspective and value. DENSITY OF LIQUID DE UTERI U M BROMIDE1 BY M. GOLDBT~ATT AND E. S. ROBINSON Los Alamos Scientific Laboratory, University of California, Lo8 Alamos, New Mexico Received Mag 81, 1966
The density of liquid deuterium bromide has been determined over a temperature range of 0 to 25". Within this temperature range, the density of liquid deuterium bromide under an atmosphere
Vol. 60
in the presence of a heated platinum catalyst. As a source of deuterium, tank stock of 99.6% deuterium content was used with no attempt a t further purification. The bromine, which was of analytical grade, was stored over PZOSto ensure that the halogen be dry. A deuterium bromide free of bromine and water was obtained by allowing the material to flow through a trap cooled by carbon dioxide and trichloroethylene and be condensed in a trap cooled by liquid nitrogen. I n order to measure the density of liquid DBr, a micropycnometer made of straight glass capillary tubing was used-much in the manner of J. M. Furter.2 The mean radius of the capillary to be used was determined by the mercury method outlined by Mann and Purdie.s The deuterium bromide gas was transferred into the pycnometer by cooling the capillary tubing with liquid nitrogen. By sealing off the tubing at a short distance above the level of the solid DBr, the vapor volume above the deuterium bromide, when it became liquid, was minimized. The pycnometer was clamped in a vertical position so that a cathetometer could be used to read both the height of the liquid column and the length of the vapor space. Mixtuw of ice and water were used to give bath temperatures ranging from 0 to 25'. The bath container was a dewar, which had been evacuated but not silvered, aIlowing the liquid height to be measured through the dewar walls. Linearity of the index marks on the capillary made it possible to tell that there was no optical distortion. After the run, the weight of the DBr waa ascertained by weighing the pycnometer before and after breaking it, and allowing the gas to escape. Then the density was calcu-
1.95C
1.90(
3 . 5 st cn 2 W
0
1.85( A-
PYCNOMETER I
0 - PYCNOMETER 2
- PYCNOMETER
3
n 1.801 I
5
I
IO
TEMPERATURE
I
I
15
20
O C .
Fig. 1.
of deuterium bromide gas can be found from the relation D = 1.961 - 5.981 X 10-St + 3.503 X 10-8t2 (1) Experimental Methods.-The synthesis of D B was ~ accorndished bv the direct union of the elements..~ Dz and Br2, (1) This work wassponsored by the U. 5. Atomia Energy Commission.
lated for each temperature from the relation
where hf is the total weight of DBr, m is the CalCUlatd weight of material in the vapor phase, H is the height of thr (2) J. M. Furter, Helu. Chim. Acta, '21, 10G6 (1938). (3) F. G.Mann and D. Purdie, J . Chern. Soc., 69,13 (1937).
I
