The adsorption of oleate on apatite was studied at pH values from 6 to 9.8 and oleate concentrations from 2 X 10"5 6to 3 X 10~4 mol Lr1. Diffuse reflectance infrared Fourier transform spectroscopy has been found to be superior to the transmission infrared technique for detecting adsorbed oleate species. Confusion in previous studies is clarified and a better understanding of the adsorption mechanism obtained. Chemisorbed oleate corresponds to a single peak at 1550 cm"1 and probably comprises one oleate ion bonding with one lattice calcium ion on the surface. Surface calcium oleate precipitate showing peaks at 1574 and 1538 cm-1 has a structure similar to that of bulk calcium oleate and probably adsorbs through ion-dipole interaction and hydrocarbon chain association. Oleic acid dimer and monomer adsorb via hydrocarbon chain association onto underlying chemisorbed oleate and correspond to a sharp peak at 1713 cm'1 and a shoulder at 1732 cm"1, respectively. Chemisorption of oleate on apatite occurred under all conditions studied and was accompanied by physical adsorption of calcium oleate precipitate and/or oleic acid monomer and/or oleic acid dimer, depending on the pH and concentration of the solution.

The interaction of ammonia with silicas prepared by a variety of methods was studied to resolve conflicting reports of the mode of ammonia adsorption and to determine the role of chlorine impurities (present in the silica) in the adsorption process. Results for the adsorption of water on silica assisted in making band assignments and competition between ammonia and water for silica adsorption sites was observed. It was concluded that dehyd;oxylated silicas contain sites which dissociateammonia to form Si-NH2 groups having infrared bands at 3540.3450. and 1550 cm-' (the surface amine HrouDs are not dis~laced bv added water).

Discrepant interpretations regarding the frequency of the C=S stretching mode are reviewed and assignments of frequencies for the vibrational modes of the xanthate group, S -0 -C~ , are presented."s-;\ band a t 1020-1070 cm-I is assigned to the C=S stretching mode. Bands a t 1200 cm-1 and 1110-1140 cm-' are ascribed to the stretching vibratiolis of the C-0-C linkage. The effects of the alkyl hydrocarbon chain length and of metal atorns (alliali metals, copper, and zinc) in displacing solne of the frequencies are recorded.

Infrared spectra of CO and CO:: adsorbed on chromia-alumina and on alumina surfaces have been determined. A band near 2200 cm-I formed by CO on both surfaces a t room temperature was due to a weak, non-activated sorption, but also contained a co~ltribution from a more strongly sorbed, activated species on the chro~nia-alurni~la. The assignment of this band was discussed in some detail. Bands in the region 1200-1800 c~n-' were co~lsiclered in terms of surface CO2-species, although in certain instances the appearance of bands a t 1750 and 1430 cm-' may have indicated carbonate ion formation.Followi~lg an infrared study of carbon ~llonoxide and carbon dioxide adsorbed on zinc oxide (1) investigations have now been extended to chromia-alumina and alumina catalysts. Previous studies of CO and COz adsorption using the infrared method with both metals and metal oxides include the well-ltnown work of Eischens and Plisltin (2, 3), correlations of vibration frequencies and number of valence electrons of CO surface species by Gardner and Petrucci (4, 5), and work on titania surfaces by Yates (6). EXPERIMENTALThe chromia-alumina catalyst was prepared by mixing chromium hydroxide and a l u n~i~l u~l l hydroxide gels which had been precipitated separately from solutiolls of the metal nitrates by ammonia. The mixed gel was then dried a t l l O o C for 24 hours and calcined a t 500' C in oxygen, resulting in a con~position of 41yo by weight of Crz03. The powder was compressed a t 40 to~ls/in? into a thin plate suitable for the infrared studies. The plate, which measured 1.3 cmX2.8 cmXO.1 mm, weighed about 100 ~n g and had a surface area of 108 m2/g. I t was mou~lted in a cell of 10-crn path length, si~nilar t o that described earlier (7). Cells of 5-mm path length were also used occasionally. Evacuation a t 300-400° C and 10-4-10-6 mm was carried out for 1-2 hours before gases were adsorbed. Spectra were then recorded on a Perkin-Elmer 21 spectrophotometer fitted with a sodillm chloride prism. The percentage transmission through the sample was 10-15yo a t 1800 cnl-I. The alumina catalyst was prepared in a manner analogous to that described for the mixed gels. RESULTSFigure l a shows the spectrum of the chromia-alumina sample, with a weak band appearing a t 1370 cm-' after heating for 1 hour in oxygen a t 320' C and outgassiilg a t rnln for 1 hour. Scattering by the sample of infrared radiation a t higher frequencies precluded ally study above 2500 cm-'.When the sample was allowed to stand in the cell a t room temperature for 24 hours a weak band appeared a t 1580 cm-I (Fig. lb). Then after further evacuation a t 350' C a broad intense band appeared a t 1530 cm-I with shoulders a t 1420 cm-I and a t about 1350 cm-I (Fig. lc). I t was only after first heating the sample in oxygen a t 300-350' C and then evacuating that the absorption in the 1300-1600 cm-I region was decreased to the level shown in Fig. la. This oxygen treatment led to an i~lte~lsification of the 1370 cm-' band.

No abstract

The adsorption of carbon monoxide onto magnesium oxide was investigated. The formation of surface carbonate groups was observed on samples prepared in high vacuum while added oxygen was necessary for carbonate formation with magnesium oxide subject to more vigorous outgassing in ultrahigh vacuum. Carbon dioxide produced similar carbonate species to that formed by carbon monoxide.

The infrared spectrum of ammonia adsorbed on porous glass a t 20' C and 150" C has been studied in the region 1450-4000 cm-l. No absorption band due to the asymmetric bending mode of ammonia was observed but in the N H stretching region, bands occurred a t 3280 cm-I, 3320 cm-', 3365 cm-I, and 3400 cm-I. The bands a t 3320 cm-1 and 3400 cm-I were easily removed by evacuation and are due to ammonia molecules hydrogen bonded through the nitrogen atom to surface hydroxyl groups. The bands a t 3280 cm-I and 3365 cm-1 were not removed by evacuation even a t 1.50' C and are due to ammonia molecules held to surface Lewis acid sites by the nitrogen lone-pair electrons. The site for this adsorption is not a surface hydroxyl group. These results are further evidence for the existence of the two adsorption sites proposed by Folman and Yates. Deuteration of the surface OH groups was easily accomplished with DzO vapor a t 300' C and the rate of hydrogen exchange between adsorbed ammonia molecules and surface OD groups was found to be rapid.

Frequency displacements and intensities are reported for the C=O stretching fundamental and first overtone in acetone, acetaldehyde, and diisopropyl ketone, for the chloroform C-H stretching fundamental, and for the acetonitrile C-C stretching fundamental, all in a variety of non-polar and polar solvents. The solvent displacement of C-C is very small (?1 K), for C-H and C=O it is to the red and of the order of 10-20 K, with C=O overtones being displaced about twice as much as the fundamentals. The Kirkwood relation between the frequency displacement and solvent dielectric constant is inadequate if the static D is used. The C=O results can be. interpreted in terms of two superimposed effects : (i) the electronic polarization of the solvent causes a frequency shift related to the solvent refractive index, and (ii) in polar solvents there is an additional shift due to solvent dipole orientation. Effect (ii) causes an added contribution to the intensity. The C-H results do not fit easily into this interpretation.