A Magnetic Investigation on Supported Molybdenum Dioxide

ride was melted in a platinum dish and oxygen was bubbled through a perforated platinum ... lets, most of which were transparent. Again it wasshown, b...
8 downloads 0 Views 1MB Size
Aug. 5, 1959

3843

A MAGNETIC INVESTIGATION ON SUPPORTED MOLYBDENUM DIOXIDE

Experiments 4-7 were carried out in open platinum dishes and produced only polycrystalline BaFe12001..Anal. Found: Ba, 12.35; Fe, 60.23; F, 0.02. Calcd. for BaFenox.: Ba, 12.36; Fe, 60.29; F, 0.00. Finally, Fe20r w a ~introduced under completely nonstoichiometric conditions into a great excess of~moltenbarium chloride. Exactly 300 g. of anhydrous barium chloride was melted in a platinum dish and oxygen was bubbled through a perforated platinum tube into the melt. At

ier and 0.1N HCI. The product consisted of lustrous platelets, most of which were transparent. Again it wasshown, by means of an X-ray diffraction powder pattern, that even under these conditions BaFelzOiQwas obtained as the only phase, which is further an indication for the high crystal energy of this particular barium ferrite. Finally, 10.00 g. of FexOs reacted in 300.00 g. of molten SrCI. under conditions identical to those described above. As before, shiny, mostly transparent platelets of SrFet2019were obtained. (Anal. Found: Sr, 8.17; Fe, 62.89. Calcd. for SrFeL20.: Sr, 8.25; Fe, 63.12.) The d-spacings of the X-ray powder diffraction pattern of BaFelpOlpwere in perfect agreement with ASTM File Card 7-276 and those of the isomorphous SrFeI2Ol9showed the same intensity variations with a come-

and wire mounted in a 114.6 mm. camera. kxpos& times to copper K-alpha (X 1.5418 A.) radiation varied between 8 and 10 hr. Aluminuni foil was placed over the film to minimize fluorescence effects. Single crystal data were obtained by precession about the c-axis in a recession camera. 0 and 1 level Dhotoernnhs were taken k i t h copper radiation and the elemeitaryhe;agonal cell dimensions were found to be ao = 5.88 A.,~co= 23.20 A. for BaFeaO,& and ao = 5.86 A,, co = 23.00 A. for SrFenO.. The space group is D& ( P 6Jmmc.) With two formula weights per unit cell, the X-ray densities are 5.31 g. cc.? for BaFe1201p and 5.15 g. CC.? for SrFeaOm in excellent agreement with the pycnometric densities of 5.22 and 4.98 g. CC.-', respectively. Domain Pattern-The particular interest of the compounds prepared by the technique described lies in the fact that they are transparent a t visible wave lengths. It must be assumed t h a t for this special form of BaFelzOlp and SrFeaO19all iron ions are 3-valent, since the presence of even small amounts of 2-valent iron would cause a significant absorption of light and make the crystal opaque. In his paper, "Direct Observation of Weiss Domains by Means of the Faradav Effect.'' Koov (ref. 4) shows oictures of intemal domain gatterns and als6 'gives 'an expianation as to how these pictures come about. Figure 1 shows a virgin state domain pattern of B single crystalline platelet of

[CONTRIBmION PROM

TAE

I'ig. 1.-~II0lil:iill p i L l t C r I I o r srl~c,20,g. SrFe,sOlq as obtaincd by the interaction of SrCh and Fe901. Transparent strontium-ferrite has not yet been reported in

more ordered form if it is observed jfter the crystal has been magnetized. With crystals of varying thicknesses, we fur-

anisotropy wit11 increasing temperature.

_ _

Acknowledgments.-I am grateful to Mr. Keith R. Babcock, Central Research Department, du Pont Experimental Station, for his assistance in the single crystal work, and to Mr. F. J. Baum of the same organization for taking the first pictures of the domains. WILMINGTON, DELAWARE

HALDOR T O P S ~RESEARC~E E LABORATORY]

A Magnetic Investigation on Supported Molybdenum Dioxide BY ERLINGR@sT' RECEIVED JANUARY 13, 1959 The magnetic susceptibilities of a series of precipitated molybdenum oxidealumina preparations have been measured in a hydrogen atmosphere after hydrogen reduction a t 44W450'. Crystalline M o a shows 3 temperature independent per gram. By dilution of the molybdenum oxide on the alumina support, the susceptibility susceptibility of f0.33 X of MOO?increases considerably. For the highest diluted compound, with an atomic ratio of Mo/AI = 1/99, the Mooz susceptibility was determined as 27 X a t 78°K.. 9.8 X 10-e a t 195°K. and 6.5 X 10-e a t 295°K. The increase in paramagnetism on dilution probably is caused by an increasing number of unpaired electrons due to the dewease in number of Mo-Mo bonds present in the crystal lattice of Moop. The susceptibility of the dispersed MOO. decreases considerably after exposure to air a t room temperature, possibly due to chemisorption of oxygen. The investigation indicates close relationships between degree of dispersion and chemisorption eficiency of the supported MOO*.

Introduction A number of transition metal oxides show an increase in magnetic susceptibility by dilution on a ( 1 ) chemical institute A. university of osio.mindern, N

~

high-area diamagnetic support, such as y-Al$Os. The usual explanation of this phenomenon is that exchange interaction covalent bonds exist between adjacent metal atoms By ~ ~ ~ ~ . in the crystal lattice.

3511

ERLING R~ST

T’ol.

s1

The sphere was connected to the magnetic balance by dilution of the oxide this interaction becomes weaker, which results in an increase in paramag- means of deKhotinsky cement, and measurements were carried out a t ’78,195 and 295’ K. The apparatus was netism. Magnetic investigations on supported evacuated during the measurements. Afterwards the oxides have been carried out mainly by Selwood samples were weighed, and the magnetism of the empty and co-workers. Selwood2 has reviewed a number containers were measured for the purpose of corrections. of such supported oxide systems. Results The magnetic susceptibility of molybdenum X-Ray powder photographs show that Moon is dioxide is about 4-0.3 X lo4 per gram. This value is much lower than expected for the Mo4+ formed in the molybdenum-rich samples by reion with its two unpaired d-electrons. Magr~Cli,~-~duction with dry hydrogen for 17-18hr. a t 440-450O. who has determined the crystal structure of Mooz, Samples containing small amounts of molybdenum found alternating short and long Mo-Mo distances oxide gave no reflections from molybdenum corresponding to the formation of pairs of molyb- compounds, possibly because of the crystallites denum atoms with the rather short interatomic of MOO*being too small. Herington and Rideal,’” distance of 2.50 A. On account of this PaulingG as well as Moseman,ll report that Moon is stabilized suggests that Mo-Mo double covalent bonds when supported on y-alumina. Accordingly, it is supposed that after the reduction also the most are formed. diluted samples studied here contain molybdenum Selwood, Hill and Boardman,’ as well as Eischens and Selwood,ahave studied the magnetic properties in the form of MOO,. ‘IVhen precipitated, anof various supported molybdenum oxide prepa- nealed and reduced in the same way as the suprations after reduction a t somewhat higher tempera- ported MoOs samples, y-A1203 showed a temperatures than that used in the present work. They ture independent susceptibility of - (0.34 f 0.02) report that the investigations failed to yield any X lo+. The susceptibility of MOOPobtained by reduction of M o 0 3in the way described above was appreciable paramagnetism. The present investigation shows that the sus- +0.33 X l o + , independent of temperature. ceptibility of b l o O ~ ,prepared by hydrogen re- For samples reduced a t 46C-470” the susceptibility duction of a precipitated mixture of MOO:, and increased to 0.4-0.5 X 10“. A small amount of A1203 a t about 450°, increases considerably by molybdenum metal could be detected in the X-ray lowering the concentration of molybdenum, a t photographs of such samples. The susceptibilities of MoO2 in the reduced, least when the oxide mixture is protected against supported samples are listed in Table I. air exposure. Experimental TABLE I Atomic Magnetic susceptibility, x a X 10’ of The supported MoOa samples used in the experiments ratio supported Moot were prepared by K . S#ndergaard of this Laboratory. &IO,’U i8’K. 195%. 295’K. Solutions containing calculated amounts of Al(NOa)s.SH*O 1/99 27 9.8 6.5 (Baker’s Analyzed) and (XH4),&lo,0~44H20(Merck, pro 8.9 5.6 2/98 24 analysis) were precipitated with concentrated ammonia. The slurry was wcll mixed and dried. After annealing for 6.4 4.1 4/96 1i 3 hr. at 500’ in the air, the samples were crushed in a mortar 3.7 2.5 8/92 8.9 and sieved. The 60-80 mesh fraction was used. Measure1.9 1.36 16/84 4.5 ments were carried out on preparations with these Mo/AI 1.2 0.94 32/68 2.6 atomic ratios: 1/99, 2/98, 4/96, 8/92, 16/84, 32/68 and 50150, 0.92 0.78 50/50 1.7 The supported molybdenum oxides mere reduced in deoxidized and thoroughly dried hydrogen for about l i hr. The reduced samples are pyrophoric and, after at 440-450° on a disk of fritted glass inside a Pyrex glass apparatus. The air was removed from the apparatus by exposure to air a t room temperature, their susceptiseveral evacuations and purgings with hydrogen. During bility decreases appreciably. This decrease is most the reduction process a slow stream of hydrogen was passed marked during the first minutes of air exposure, through the sample. Afterwards, the reduced samples and after a few days the susceptibility becomes conwere transferred t o a thin-walled Pyrex glass sphere, with a dismetcr of about 7 mm., connected t o the apparatus stant. Due to water adsorption the weight of the samples increased 9 to 17%. The room temperathrough a capillary. The sphere was sealed by melting off the capillary. ture susceptibilities of Moo2, as measured in H2, The magnetic nieasuremcnts were carried out by t h e and also after air exposure, are plotted versus Faraday method, and the balance was of the Sucksmith type. T h e magnet and balance were designed by 0. Gram composition in Fig. 1. The latter values are corJeppesen of this Laboratory. Calibrations were made rected for the water content using a susceptiagainst ( XIi4)pFe(S0,s)y6H?0as recommended by S e l w o ~ d . ~bility of -0.80 X per gram adsorbed water. A t a pole piece distance of 25 mm., the product N X dH/& This rather low value was found for water adsorbed has its maximum value of about 50 X lo6 oersteds*/cm. and on pure y-alumina. I t corresponds, however, a field strength H of about 11,000 oersteds. ( 2 ) P. n’. Selwood, “Advances in Catalysis,” Vol. 111, Academic Press, Inc., T e w Y o r k , N. Y.,1991, pp, 27-106. ( 3 ) A . Magn6Ii. A r k i o K e m i , M i ~ r e r n l (hi., . 8 2 4 , 1 (1946). (4) A. Nagneli and G. Andersson, .A