KELSOB. MORRISAND PRESSL. ROBIXSON
1194
Electrical Conductance and Density in Molten Molybdenum(VI) Oxide-Alkali (Lithium, Potassium) Molybdate Systems'
by Kelso B. Morris and Press L. Robinson Department of Chemistry, Howard Uninersity, Washington, D . C. 80001
(Received M a y 28, 196s)
Electrical conductance and density have been determined as functions of temperature for the two fused binary molybdate systems, Kz,l/Io04-1100~ and LiZ11o04-MoO3. Data for the two properties in both systems cover temperatures which range up to at least 100' above the melting point for each composition. For the single components, the ranges for specific conductances, in ohms cm.-', are 1.211 a t 931.0' to 1.302 a t 997.8' for K2R!Io04; 4.017 at 790.6" to 6.761 a t 939.0' for Li211004; and 0.685 a t 823.8' to 0.719 a t 820.9' for are 2.479 at 935.0' to 2.450 at 988.3' for nloOa. Ranges for the densities, in g. K2Mo04; 2.888 at 781.2' to 2.802 a t 963.4" for Li'2i\1004; and 3.206 at 828.2' to 3.083 a t 918.0' for R1003. When conductances for the first system, KzMoO4-MgO3, are compared at corresponding temperatures that represent a fixed number of degrees above the melting points, resemblance is observed between the conductance-composition diagram and the phase diagram for this system. In plots of either specific conductances or equivalent conductances, at corresponding temperatures, us. composition, two conductance minima and a single conductance maximum correspond, respectively, to the eutectic compositions and compound composition in the phase diagram for this system. -4 conductance-composition plot for the second system, Li2;\1004-RIo03, in terms of the more familiar polytherms is also similar to the phase diagram for this system. Although the phase diagram for this system shows peritectic reaction and no compound formation, the composition for the single minimum or true eutectic in the system is the same as that for the conductance maxima in the polytherms, The two kinds of isotherms together with references to the observations of others suggest strongly that correlations may exist between conductance-composition diagrams and the phase diagrams for many molten binary systems. An explanation for the variation of conductance with compositioii in both systems is based on the nature and relative contributions of the constituent ionic species suggested by the phase diagram. Average activation energies, based on specific conductances, for the single substances and Naz31004 a t corresponding temperatures that are 100' above the melting points decrease in the order Li2Mo04> Sa2MoO4> K2J1004 > 1 \ / 1 0 0 3 or L L I o O ~ +R~I~O O ~ - ~ .
Introduction I n an earlier paper,* electrical conductance and density data were reported for the molten system, molybdenum (VI) oxide-sodium molybdate. The present report is concerned with measurements of the same properties, with superior scientific equipment and vastly improved experimental procedures, for two additional molten systems consisting of molybdenum(VI) oxide and the molybdates of lithium and potasThe Journal of Physical Chemistry
sium. Density data and conductance data appear not to have been obtained previously for these two systems.
(1) This research was supported by National Science Foundation Grant No. 9487 to Howard University. I t is based in part on the dissertation submitted by Press L. Robinson to the Graduate School, Howard University, in partial fulfillment of the requirements for the degree, Doctor of Philosophy. (2) K. B. Morris, M. I. Cook, C. Z. Sykes. and M . B. Templeman. J . Am. Chem. Soc., 77, 851 (1955).
1195
ELECTRICAL CONDUCTANCE AND DENSITYIN BINARY ~ ~ O L Y B D A TSYSTEMS E
However, Jaeger3 has reported density data for pure KJMo04.
Experimental Chemicals. All inolybdenum-containing chemicals were obtained on special order from the S. ’VV.Shattuck Co., Denver, Colo., as reagent-quality anhydrous powders. A wet method of chemical analysis based on total molybdenum revealed these purities : MOOS, 99.5%; LiJlo04, 99.7y0; and KAb4004, 99.7%,. Melting points, as determined by the method of thermal analysis for these substances, were in agreement with literature values. General qualitative spectrographjc analysis of two cheniicals a t the National Bureau of Standards revealed impurities to be less than O.Olyo for M o o S (La,b. Xo. 5.02-34726) and less than 0.1% for K21\1004(Lab. No. 5.02-34727). Electrical Conductan,ce. The experimental details employed were essentially the same as described in a recent report4 on electrical conductance in another fused molybdate system. For the resistance measurements, the assembly consisted of a Leeds and Northrup Co. precision Jones conductivity bridge (Catalog No. 4666), a Hewlett-Packard Co. oscillator (20 C.P.S. to 20 k.c., Model 201-C), and a General Radio Co. tuned amplifier and null detector (Type 11232-A). All leads were shielded. The dip-type conductance cell of clear fused quartz was identical in basic design with that described by Van Artsda1en.j Calibration of each cell was carried out by using the technique suggested by 0thers.j Platinum disk electrodes, each having a 3-mm. length of platinum-rhodium alloy wire (B. & S. gage 24), welded to the underside of the disk, rested on the shelf a t the top of the capillary tubing in each arm of the quartz cell. The electrode leads, also of platinum-rhodium alloy wire (B. & S. gage 24), were covered with several refractory mullite insulators up to the point where they were joined to short copper leads just outside the furnace chamber. These insulators, each about 5 em. in length, served to keep the electrodes seated firmly and to minimize the mutual inductance that otherwise might be caused by the close proximity to each other of the electrode leads. The calibrated conductance cells, with cell constants of the order 149-267 cm.-I, were discarded after being used in no more than two runs because cumulative attack of the quartz by the molybda1,e melts was then significant. All resistance nieasurements were made a t a frequency of 2000 cycles. !In making a measurement, the power supply to the furnace was shut off temporarily in order to prevent the 60C.P.S. current to the furnace from interfering wit,h bridge balance. The furnace insulation was sufficient
to maintain constant temperature during this short interval. Density. Measurements of density and the calibration technique for the specially designed density bob, machined from a single solid platinum cylinder and with a weight of the order 27.0000g., were essentially the same as described by J a m 6 The diameter of the platinum-rhodium alloy wire by means of which the bob was suspended from one arm of a Christian Becker chainomatic balance was 0.020 in. except a t the point where the wire contacted the melt surface. For this point, a short length of platinum-rhodium alloy wire (of 0.01-in. diameter) was used between the large wire and the stem of the density bob. Therefore, surface tension corrections were negligible. Temperature. A Marshall tubular testing furnace (6.35-cm. i.d., 30.84-cm. o.d., 40.70 em. long), wound with platinum-rhodium alloy wire, was used for the heating and melting of all samples. The furnace temperature was controlled to 0.5’ by means of a millivoltmeter pyrometer controller operating in conjunction with a Pt-Pt 137, Rh thermocouple located directly beneath the heating element and running the full length of the furnace chamber. A pIatinum-ZO% rhodium alloy crucible of 70-mI. capacity was used as container for all samples. Preliminary drying to constant weight as well as the melting, heating, and cooling of all samples were carried out in an atmosphere of pure argon gas. Temperatures of the melts were measured with an NBS-calibrated Pt-Pt 10% Rh thermocouple by means of a Leeds and Northrup Co. Type E(-3 pot)entiometer (Catalog Xo. 7553-5) while the lower end of the thin-walled thermocouple protection tube was approximately 1 mm. (as determined by a B & S super vernier caliper, accurate to 0.001 in., in positioning the crucible and sample with a CencoLerner Lab-Jack) abovk the surface of the melt. One important aspect of the research was the development of a satisfactory procedure for determining the composition of the melts. Since the vapor pressure of pure molybdenum (VI) oxide is significantly high in the vicinity of its melting point, the vapor pressures of mixtures of this component and alkali molybdates are appreciable. A qualitative indication of the magnitude of this partial vapor pressure of the Mo03 is obtained by observing the extent to which there is ~
~~
“International Critical Tables,” Vol. 3, McGraw-Hill Book Co., New York, S . Y . , 1928, p. 24. (4) K. B. Morris, N . McNair, and G. Koops, J . Chem. Eng. Data, (3) F.
M. Jaeger, 2. anorg. allgem. chem., 101,
16 (1917);
7, 224 (1962). ( 5 ) E. K. Van -4rtsdalen and I. S. Yaffe, J . Phys. Chem., 59, 118 (1955). (6) G. J.
Jans and AI. R. Lorens, Rev. Sei. Instr., 31, 18 (1960).
V O ~ U68, ~ Number E 6 May, 1964
1196
KELSOB.
smoking of the rnclts. The vapor pressures of the pure, nioltori alkali niolyhdates arc t1c:gligiblo for the ternpcratjuro ranges cniployed. Thwefore, the coniposition of each sample changes slightly during tho hcating and cooling process because of the loss by volatilization of small amounts of 1100,. Aloroover, these changes in composition were found to deperitl principally upon the pcrccntagc of AIoOa in cadi mixture, tho tcnipcraturc, and thc duration of the heating aitd cooling period. Now, if chemical analyses of mixtures such as these arc to be ineanitigful, they tnust be made of tiiclt portions taketi a t thc prccisc tinies the properties, electrical oonductancc and dcnsity arc measured. Howevcr, a sanipling process of this kind followcd by rapid quenching was not feasible for the cxperimcntal conditions employed. A decision was madc to subject the
M O R R I S ANI) I’RESS
1,.
I%OBINSON
cooled mixturcs to chcniical analysis for total molybdcr m n (conversion of MoOa to molybdate ion, precipitation of tiiolybdatc with a-benzoin oxiiiic, and subscquont ignition of the precipitate under appropriate conditions to the oxide which is then weighed). This well-known analytical procedure was found to be satisfactory for unheated samples and for cooled melts of thc sanic initial cotnppsitiori. Subscqucnt investigation revealed that csscntially the S a m infortilation on final compositions of the inixtures could be obtuincd much niore conveniently by carrying out a series of weiglit-losson-heating experinicnts. In usitig this tcchniquc, weighed samples of a given mixturc were heated under the same conditions that prevailed whcn the propertics were actually being measured. At certain titrie intervals, crucibles containing thc riiolts were withdrawn
Table I : Specific Conductmccs for the System KJloOa--AIoOa Colnpn..
mole 5% MOOS
0.00
14.13
60.20
K, 1,
=e.
mhos cm.
931.0 938.7 949.0 955.8 964.7 972.6 983.0 990.6 997.8
1.211 I ,233 1,248 1,266 1.271 1.280 1.290 1.299 1.302
876.8 883.0 897 2 902.4 910.0 917.8 926. H 931.3 949. 5 964.7
0,902 1.014 1.020 1.028 1.041 1.0,?6 1,071 1.083 1.115 1.144
664.0 673.2 682.1 689. I 701 .o 707.2 719.3 729.0 739.5 752.0 764.3 771.7 781.0 780.0 794. (i
0.473 0.488 0.503 0.512 0.532 0.546 0.565 0.580 0,597 0.619 0 . 639 0.647 0,660 0 .6GR 0,672
T h e Journal of Physical Chemistry
Compn., mole ?&
1,
h100S
OC.
mhoe c m . -1
14.13
975.4 985.7 988.4 9!)4.7
1.161 1.170 1.194 1.215
30.13
882.2 830.8 838.0 845.8 853.7 862.7 871.4 877.2 891 .o 900.0 914.0 928.6 933.8
0.882 0.895 0.908 0.920 0.933 0.945 0.958 0.968 0.989 1.002 1.023 1,037 1.044
52.44
583.3 587.7
0.328 0.334
68.64
689.0 699.5 703.6 718.1 727.5 738.0 754.0 767.8 771.4 776.5 782.8
0.540 0.560 0.576 0.595 0.616 0 .ti35 0.660 0.677 0.687 0.696 0.708
K,
Compn.,
76
K,
t, OC.
mhos cm.7
52.44
5!13.G 598.5 604.0 609.7 616.1 626.0 639.8 649.8 657, 6 667.1 674.5 679.1 682.8 688.7 697.8 704.6
0.345 0.354 0.368 0.377 0.391 0.407 0.427 0.443 0.455 0.465 0.473 0,480 0.486 0.493 0,502 0.504
60.20
642 . 0 649,6 656. 6
0.434 0.450 0.463
77.35
765.3 775.0 786.3 798.0 809.1 818.3 826.8 836.8 845.8
0.653 0.677 0.699 0.719 0.733 0.747 0.760 0.771 0.778
mole
MOOS
1197
ELECTKICAL CONDUCTANCE A N D DENSITY IN BINARY MOLYBDATE SYSTEMS
Table I
(Continued)
Compn..
yo
K,
1, OC.
mhos cin. - 1
68.64
654.4 662.0 669.7 679.2
0.473 0.486 0.499 0.521
81.35
738.0 749.8 755.6a 768.7” 774.2 783. l o 793.0 803.9
0,589 0.611 0,625 0.643 0,661 0 677 0,693 0.706
700.2 771. 0 789.0 799.2 810.2 818. Oa 827.8 838.0 845.8 852.2 859,O
0.613 0 030 0.0h which investigators have been corifrontcd is the selcction of a suitablc temperature at which properties of various melts can be compared. For sonic substanccs, either the boiling point, or some teinpcraturo related to the critical teiiipcrature has bceii found uscful as a “corresponding tctriperature.” I t is possible that at, some tinic in the future it may bc found nieaningful to compare physical properties of substanccs at a tcmperature that is the melting point plus some fraction of thc liquid range for each pure substance under a
pressure of 1 atm. Howcvcr, boiling point data for fuscd salts arc cithcr ‘ unrcliablc or iiot available. Rcscarchcrs in the area of fused salts have used tw.o general approaches iii an effort to arrive a t a satisfactory “correspoiidiiig teinperature.” Soiiie investigators’ have used coiiduct~ance-coiiipositiori plots for a givcii systeni at several arbitrarily choscn tmiperatures. ’l’hesc polytherins have bccn iiseful in an interpretation of the data. This is doiict in I’ig 2 of t’he present report on the molten system, I,iJIo041 0 0 3 . Still othersR have used conductaiicc-coniposition plots at, an arbitrarily defined “corrospoiiding tJcnipcraturc.” Van Artsdalcn,8 iii his studies 011 nioltcn salt syst(iins, coinparcd the proportics of riiclts at equal fractions, e, of the mcltiiig point on tho Kclviri scale, where 8 = Y°K./T,,LoI
