Magnetic Susceptibility of the Lithium Tungsten Bronzes 13Y IAWRENCE E , coxROlr )lND hI. J. S ~ E N K O KBCLIVEU FEBRUARY 9, 1932 1,ithiuni taiigsteii bronzcs, differing in lithium coiicentration, have been prepared and their magnetic nionients dcterrnincd by thc Gouy method. The molar susceptibilities are very low7 and actually become diamagnetic as the concentration of lithium in the bronze is reduced. The results are consistent with magnetic susceptibilities calculated for an electron gas detcrininrd i n thii vrork is lower thaii t h a t previously reported. model The molar susceptibility of M70?
The term tungsten bronze is applied to the uonstoichiometric compounds of general formula MzWO, (M is an alkali metal, x is less than unity) formed by the reduction of alkali tungstates. Hagg' and Straumanis3 have considered these hronzes to be solutions of tungsten(S'1j oxide in the compound MWOs. Because of the high conductivity2 and the low magnetic nioinents3, i observed in these bronzes, it has been suggested" that they can more properly be considered as solid solutions of alkali metal in WOS and that the properties observed can be explained on the basis of an electron gas model such as is accepted for metals. In this investigation we have prepared a series of lithium tungsten bronzes in which the concentration of lithium in WOS is varied over the widest limits possible. The magnetic moments have been ineasured and compared with those expected for these two models.
water.
Reaction began immediately and was completed
by warming to 100" for 2 hours. The water was evaporated off a t 80" and the white powder-cake product dried at 110'.
Analysis of the Lithium Tungsten Bronze, Lio.36WO~.-?'he analysis was complicated by the diffculty of getting the brorizc into solution. Fusion with ammonium sulfatesulfuric acid mixture, which has been rcported,5 did not cornc thc bronze. The ammonium su1fate~urc attacked only the surface. Sincc rere tedious, the method used in this work was fusion with a sodium nitrate-sodium carbonate mixture despite the fact t h a t the fusion introduced sodium ion into the analysis solution. Lithium was determined with a Beckman model DU spectrophotometer with Beckman flame photometer attachment. 'l'he instrument was standardized with a lithium solution containing all the other constituents of t h e analysis solution in the approximate proportions present in the analysis solution. The 670.8 mfi line of lithium was used for the measurements. the intensity being measured on the 600-1000 mp phototube. The precision was 9.2% in terms of the lithium coiitc'nt of the solution. Tungsten was determined by precipitatioii of the tungstic acid--tannin complex in the prescnce of cinchonine,6 folExperimental lowed by ignition to !TO$. Magnetic Measurements.--Magnetic susceptibilities of Preparation of Lithium Tungsten Bronzes. -The most concentrated bronze obtainable ( t h a t is, the bronze with the the bronzes were determined by the Gouy method using a highest. Li t o \$Q; wtioj, Lio.36IVO3, was prcpareti by elec- senii-niicro balance sensitive to 0.05 mg. Current for thc trolytic retluctioii o l a fuscd I : 1 IiiolRr inixttire of lithium air-cooled clectromagiict was supplied by storage batteries tuilgstate and tungstcn(\'l i oxide. T h e electrolysis was a i i d the inagiietic ficld determined from the potential drop mrricd out in a porcelain crucible with a ttingstcn cathode across ; I fixcd mangatiin wire resistance in series with the coils. The field was calibrated with a 30% nickel chloride iirid a platinum auotle under the following conditions: applied voltagc, 6 volts; current, 2 t o 4 ainpci-es, temp. 800 i solutioii made up from J. T. Baker "Special Low Cobalt" 20'; cathode current density, 0.6 to 0.13 amp. pcr square SiC12.6H20. The ficld calibration was reproducible t o inillirnctcr. The bronze was deposited on the cathode as better than lyo over the range 4000-8000 gauss. It wii,, aggregates of blue-black crystals. The utireacted material coiifirnied by determining the proton resonance frequency tvith ii coil containing 0.05 -If M n S 0 4 solution as one com\\-ai leachcd out with boiling water, followed successively ponent of a regenerative oscillator. I n employing the pro\\-ith hot dilute aqueous ammonia, hot concentrated HCI, ton resonaiice technique,7 the magnetic field was modulated \vah .ipproxiniatcly inrersely proportional to the current tleri- coil indicated that the ficld was homogeneous over a distancc qity used. The largest crystcils obtaiiied w r e 0.1 r n i n . 01: of about *5 min. from the center of t h e gap perpendicular to : c i i d d o n g the magnet axis. For precision, two cross-hair
