Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Reductive precipitation of molybdenum oxides for recovery of molybdenum from hypochlorite leach solutions Reid, Duncan Craig 1979-12-31

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


831-UBC_1979_A7 R45.pdf [ 3.62MB ]
JSON: 831-1.0078731.json
JSON-LD: 831-1.0078731-ld.json
RDF/XML (Pretty): 831-1.0078731-rdf.xml
RDF/JSON: 831-1.0078731-rdf.json
Turtle: 831-1.0078731-turtle.txt
N-Triples: 831-1.0078731-rdf-ntriples.txt
Original Record: 831-1.0078731-source.json
Full Text

Full Text

REDUCTIVE PRECIPITATION OF MOLYBDENUM OXIDES FOR RECOVERY OF MOLYBDENUM FROM HYPOCHLORITE LEACH SOLUTIONS by DUNCAN CRAIG REID B.A.Sc, The University of British Columbia, 1972 M.L.S., The University of British Columbia, 1974 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Metallurgical Engineering) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA September 1979 (c) Duncan Craig Reid, 1979 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Metallurgy  The University of British Columbia 2075 wesbrook Place Vancouver, Canada V6T 1W5 October 5,1979 Date - (ii) -ABSTRACT The feasibility of reductive precipitation of molybdenum oxides as the molybdenum recovery stage of a hypochlorite leach of Cu-Mo rougher concentrates has been investigated. Hydrogen gas at elevated temperature and pressure and hydrazine at moderate temperature and atmospheric pressure were used as reductants. Reduction was performed on solutions containing 5 to 17 g/1 Mo as sodium molybdate. Hydrogen reduction was successful only in the presence of a Pt catalyst, temperature = 200°C, pressure = 30 atm of H. , and initial acidification to pH = 2. Ten hours was required to obtain 90% recovery of molybdenum as MoC^. Reduction with hydrazine yielded an MoO(OH) precipitate with 90% recovery obtained in 40 minutes at 50°C, pH = 4.5, and initial mole ratio of hydrazine to molybdenum of 4:1. Precipitation under the same conditions in the presence of 3 M NaCl gave only 70% recovery in 4 hours and the precipitate contained 3.3% sodium. The effect of NaCl is explained in terms of stabilization of mixed valent ionic molybdenum species in the presence of NaCl. - (iii) -TABLE OF CONTENTS Page ABSTRACT ii TABLE OF CONTENTS iiLIST OF TABLES iv LIST OF FIGURES v ACKNOWLEDGEMENTS vi 1. INTRODUCTION 1 2. LITERATURE REVIEW 11 2.1 Aqueous Chemistry of Molybdenum 1VI 2.1.1 Mo 12.1.2 Reduced Species and Precipitates 13 Molybdenum Blue 14 MoV 18 MoIV 21 Mo111 2 2.1.3 Summary  3 2.2 Reduction with Hydrogen and Carbon Monoxide 25 2.3 Reduction with Hydrazine 33 2.4 Reduction with S02 and H2S 42.5 Summary 44 , 3. SCOPE OF PRESENT WORK 45 4. EXPERIMENTAL 46 5. RESULTS 51 5.1 Hydrogen Reduction 55.2 Reduction with Hydrazine 54 6. DISCUSSION 77. CONCLUSION 80 8. REFERENCES 2 - (iv) -LIST OF TABLES Comparison of leach liquor solutions for molybdenum recovery Technical grade MoO^ specifications for Endako Mines Rate constants obtained for different values of initial hydrazine to molybdenum role ratio - (v) -LIST OF FIGURES Page 1. Proposed flowsheets for hypochlorite leaching of molybdenite containing concentrates. 3 2. Principal operations and products in processing of molybdenite concentrates. 7 3. Potential-pH equilibrium diagram for the system Mo-H 0 at 25°C. 8 4. Predominance area diagram for Mo^"*" in 3 N NaCl. 12 5. Reduction of ammonium paramolybdate solutions by hydrogen in the presence of a colloidal palladium catalyst. 26 6. Reduction of sodium molybdate solutions by hydrogen at 200°C and 40 atm H . 28 7. Effect of various catalysts, initial pH, and hydrogen pressure on hydrogen reduction of sodium molybdate solutions at 200°C. 30 8. Reduction of 17 g/1 sodium molybdate solution by hydrogen at 200°C and 30 atm in the presence of Pt clad niobium mesh. 52 9. Effect of sampling technique on concentration of molybdenum remaining in solution vs time at 50°C with pH = 4.5, initial molybdenum concentration -5 g/1, and initial hydrazine to molybdenum mole ratio -4:1. 55 10. Comparison of results obtained by direct filtration of slurry samples with those obtained by previous dilution of samples with an equal volume of cold water. 57 11. Distribution of molybdenum between solution and precipitates as a function of time at 50°C with pH = 4.5, initial molyb denum concentration -5 g/1, and initial hydrazine to molyb denum mole ratio s4:l. 58 12. Distribution of molybdenum between solution and precipitates as a function of time at 50°C with pH = 4.5, initial molyb denum concentration -5 g/1, and initial mole ratio of hydrazine to molybdenum =2:1. 59 - (vi) -LIST OF FIGURES (Continued) 13. Effect of temperature on rate of precipitation for pH = 4.5 initial molybdenum concentration -5 g/1, and initial mole ratio of hydrazine to molybdenum -4:1. 14. Effect of initial mole ratio of hydrazine to molybdenum on rate of precipitation at 50°C with pH = 4.5 and initial concentration of molybdenum -5 g/1. 15. Effect of 3 M NaCl on rate of precipitation at 50°C for pH = 4.5, initial molybdenum concentration -5 g/1, and i initial mole ratio of hydrazine to molybdenum =4:1. 16. Effect of addition of 0.55 g/1 Cu as copper sulfate on the rate of precipitation of molybdenum at 50°C for pH = 4.5, initial molybdenum concentration =5 g/1, and initial hydrazine to molybdenum mole ratio -4:1. 17. 1.5 order in molybdenum plots for T = 50°C, pH = 4.5, and initial molybdenum concentration =5 g/1. 18. Order in N^H^ based on 1.5 order in molybdenum. 19. Effect of temperature on rate assuming 1.5 order in molybde num for pH = 4.5, initial molybdenum concentration -4 g/1, and initial mole ratio of hydrazine to molybdenum -4:1. 20. Arrhenius plot based on 1.5 order in molybdenum. 21. Thermogravimetric weight loss curve for brown precipitate produced by reduction with hydrazine. 22. Concentrations of hydrazine and molybdenum remaining in solution vs time at 50°C with pH = 4.5, initial molybdenum concentration =5 g/1, and initial hydrazine to molybdenum mole ratio -4:1. - (vii);-ACKNOWLEDGEMENT I would like to express my gratitude to Dr. Ian H. Warren for his enthusiasm and suggestions throughout the course of this work. I am also grateful to the Steel Company of Canada for support in the form of a Stelco Research Fellowship. Last but not least, I would like to express my appreciation for the help and advice I received from the Faculty, technical staff, and fellow students in the Department of Metallurgy. - 1 -„1. INTRODUCTION Conventional technology for the production of M0S2 CONCENTRATES as a byproduct from processing of porphyry copper ores is based on a bulk or rougher flotation for the recovery of a copper sulfide-molybdenum sulfide concentrate. This is followed by differential flotation to pro duce separate copper and molybdenum sulfide concentrates. The molybdenum concentrate is then cleaned by further flotation steps to obtain a market able M0S2 product. According to data published by Sutulov"*" for fifty flotation plants treating such ores the mean recovery of molybdenum in the copper rougher concentrate is 64% while the mean overall recovery is only 2 50%. Warren et al. have shown that selective leaching of molybdenum from rougher concentrates could be an economically feavourable alternative to further flotation steps if the molybdenum recovery from the rougher con centrates approached 100% and if the residue was suitable for production of a copper concentrate by flotation. They proposed a sodium hypochlorite leach for treatment of rougher concentrates that involved on-site electro lytic generation of sodium hypochlorite using commercially available hypo chlorite cells. A similar approach has been proposed by workers at the 3—8 U.S. Bureau of Mines (USBM). In the USBM process hypochlorite leaching would be applied to the off-grade molybdenum sulfide concentrate produced by differential flotation of the initial orugher concentrate. This process, then, would replace only the cleaning stages of production of an MoS~ - 2 -concentrate while the process proposed by Warren et al. would also replace the differential flotation step. A further difference between the two proposals is that the USBM process involves generation of sodium hypochlor ite "in-situ" by electrolysis of a brine slurry of the concentrate as it is pumped through a specially designed electrolytic cell. The USBM process has been piloted on concentrates containing 16 to 35% Mo and 6 to 15% Cu. Warren et al. have designed their process to treat concentrates containing approximately 0.3% Mo and 12% Cu. The basic flow sheets for both processes are compared in figure 1. Both pro cesses use Na^CO^ for pH adjustment. Warren et al. propose to carry out leaching at pH 9 while the USBM process operates at pH 5.5 to 7. The leach temperature for both processes is 45 - 50° C. The stoichiometry of the leach can be represented by the equation: 2- - 2-9 OCl + MoS2 + 60H — •+ Mo04 + 9C1 + 2S04 + 3^0 which shows that 1 kg of NaOCl should be consumed to bring 0.143 kg of Mo into solution. According to Warren et al. modern hypochlorite generators require 3.52 kwh per kg NaOCl. It follows that in their process energy g consumption should be 24.5 kwh per kg Mo. In a recent pilot plant study the USBM process required 19.8 to 28.6 kwh per kg Mo to treat concentrates containing 16 to 35% Mo and 6 to 15% Cu. The molybdenum recovery decreased from 93 to 97% to only 75% when the copper content increased to 15% and the power consumption correspondingly increased to the 28.6 kwh value. The decrease in molybdenum recovery was attributed to the formation of insolu ble copper molybdates. Current research in this laboratory has shown, however, that the effect of copper minerals on the molybdenum recovery is Cu-Mo ROUGHER CONCENTRATE residue to copper flotat ion copper product Mo product •< No2S04 LEACH I No CO, < 2 3 COPPER REMOVAL NaOCl recycle ACIDIFICATION H2S04 Mo RECOVERY EVAPORATION NaOCl REGENERATION WARREN et al Figure 1: Proposed flowsheets for hypochlorite OFF-GRADE MOLYBDENITE CONCENTRATE Na2C03 brin e recycle Na2C03 ELECTRO-OX I DATION SO. ACI DI FICATION SULFATE REMOVAL Na2S04 SOLVENT EXT RACTION CARBON ADSORPTION CRYSTALLIZATION f (NH4)6Mo7024 - 4 H20 USBM PROCESS leaching of molybdenite containing concentrates. RESIDUE j to waste NH4Re04 solution - 4 -more complex than this. At the pH of the USBM leach, for example, it is likely that significant decomposition of NaOCl to NaClO^ occurs. Since NaClO^ does not oxidize MoS^ at this pH such decomposition reduces the energy efficiency of the leach. The use of carbonate for pH control has also been 9 shown to be an important factor in increasing molybdenum recovery. Research in this area is continuing. On the basis of the available in formation, then, it seems that the process proposed by Warren et al. offers a higher recovery of molybdenum with a better leaching energy efficiency than the USBM process. Warren et al. calculated the approximate composition of the leach solution that would result on application of their process to a rougher concentrate containing 0.3% Mo. Table I compares this with the composition resulting from application of the USBM process to a molybdenite concentrate. It can be seen that the two solutions are similar. It would thus be possible to apply the USBM solvent extraction-carbon adsorption sequence for recovery of molybdenum and rhenium to Warren et al.'s process. A disadvantage of the USBM procedure however, is the necessity to acidify 4 the solution to pH< 2 for molybdenum extraction by a tertiary amine. A process capable of recovering molybdenum from alkaline solutions would be preferable for a hypochlorite leach operating at pH = 9. Alternative means of molybdenum recovery have been suggested. These include reduction of molybdate solutions with hydrogen at elevated tempera ture and pressure to produce insoluble MoO^® ^ and reduction with iron followed by neutralization to precipitate MoCOH)^.''"6 As described in the literature, however, both of the processes operate in acid solution. The feasibility of alternative routes to molybdenum recovery from - 5 -Warren et al. USBM H20 1100 kg 900 kg NaCl 170 kg 100 kg Na2S04 68 kg 37 kg Na2Mo04 10 kg 4 - 43 kg Cu variable NaOCl„ 3 nil 5 - 13 kg NaCIO „ 4 5 - 8 kg .14 - .72 kg Re .001 - .050 kg pH 9 5-7 Table I Comparison of leach liquor solutions for molybdenum recovery - 6 -leach solutions depends on the economics of the unit operations involved and on the marketability of the resulting product. In addition the ability of a given process to recover rhenium could be a deciding factor in cases where the concentrate to be treated contained a significant amount of rhenium. Conventionally almost all molybdenum sulfide concentrates are roasted to produce a technical grade MoO^. The principal use of this product is the production of calcium molybdate, molybdic oxide briquets, or ferromolybdenum for alloy steel production. Technical grade MoO^ can also be further purified by sublimation to yield pure MoO^ or by hydro-metallurgical means to produce ammonium molybdate or sodium molybdate (Figure 2). A hydrometallurgical treatment of MoS^ containing concentrates should at least be capable of producing a product of equal quality to :techni-cal grade MoO^. By way of example Table II shows technical grade MoO^ specifications for Endako Mines Ltd. Hydrometallurgical production of a molybdenum oxide, sodium molybdate, or ammonium molybdate product of high purity directly from an off-grade or rougher concentrate could, however, eliminate not only roasting but also sublimation or other subsequent purification steps. The ability to produce such a product could influence the economic favourability of a hydrometallurgical route. The potential-pH diagram for the molybdenum water system published 18 by Pourbaix (Figure 3) is relatively simple and indicates that reductive precipitation of MoO^ could be possible over a wide range of pH. Several reducing agents might be considered. These include E^t SC^, H^S, CO, iron and hydrazine. The gaseous reagents and hydrazine are attractive because they offer the possibility of producing a high purity product. MoS. concentrate 2 i ROAST T _*-MoS2 I ubri cants t—~ calcium molybdate \ oxide brique t s technical grade Moo^ r ferro-molybdenu m DISSOLUTION 8 CRYSTALLIZATION ; sodium molybdate } f ? : } ceramics fertilizer chemicals direct addition to ste el SUBLIMATION pure MoOj DISSOLUTION 8 CRYSTALLIZATION i ammonium molybdate pigments chemicals coatings 1 T chemical catalys reagent REDUCTION Fiqure 2; molybdenum metal Principal operations and products in processing of molybdenite concentrates. - 8 -- 9 -OXIDE Range % Typical % Guaranteed % Mo 57.0 - 62.0 59.0 57.0 min. Cu 0.05 - 0.10 0.075 0.10 max. S 0.03 - 0.10 0.06 0.10 max. P — 0.01 0.05 max. Pb 0.015 - 0.050 0.025 0.05 max. Bi 0.02 - 0.04 0.030 0.05 max. wo3 0.030 0.16 max. sio2 5.0 - 15.0 8.0 15.0 max. Fe 0.020 - 0.45 0.35 — CaO 0.050 - 0.120 0.07 — OXIDE BRIQUETTES (PITCH) Range % Typical g, "o Guaranteed % Mo 50.6 - 54.0 53.0 51.6 min. C 10.0 - 15.0 11.00 12.0 approx. Cu 0.05 - 0.15 0.075 0.15 max. S 0.09 - 0.15 0.11 0.15 max. P — 0.01 0.05 max. Bi 0.02 - 0.04 0.03 — Fe 0.20 - 0.40 0.29 — Pb 0.015 - 0.050 0.025 — Table II Technical grade Mo0_ specifications for Endako Mines - 10 -Despite the apparent simplicity of the Pourbaix diagram, however, the aqueous chemistry of molybdenum is complex. There is, in fact, a rich chemistry of polymeric molybdates and reduced molybdenum species that has resulted in an extensive and often contradictory literature. A review of this literature is necessary as a basis for considering reductive precipitation as a means of molybdenum recovery from alkaline solutions. - 11 -2. LITERATURE REVIEW  2.1 Aqueous Chemistry of Molybdenum VI 2.1.1 Mo VI A hypochlorite leach of MoS^ produces Mo in solution which for 2-pH's greater than about six exists as the monomeric molybdate ion MoO^ On acidification, however, molybdate ions polymerize consuming H+ to produce isopolymolybdates such as the paramolybdate ion Mo^O ^ . Several isopolymolybdates have been reported and the literature has been reviewed 19 by Sasaki and Sillen. They characterized isopolymolybdates in terms of their acidity, z, defined as moles H+ bound per mole Mo. Thus paramolyb date formed according to the equation, 8H+ + 7Mo042~ = H8(Mo04)?6~ = Mo O^6", can be represented by z = 8/7 = 1.14. It is generally accepted that paramolybdate is the first species to form on acidification of molybdate solutions and that it is the only 20-22 isopolyanion, or at least by far predominant, up to z = 1.14 (pH - 5). The nature of the polymolybdates formed on further acidification must be regarded as uncertain since there is some disagreement in the 19 23-25 literature. ' Figure 4 shows a predominance area diagram constructed 22 19 by Baes who accepted Sasaki and Sillen's sequence of isopolymolybdates. For acidification up to z = 1.5 it has been shown that isopolymolyb-date equilibria are almost instantaneous for moderate molybdenum concentra tions. For z S- 1.5, however, where polymerization proceeds beyond the 24 heptamer or octamer stage, Aveston et al. noted that the equilibria are only slowly established. - 13 -The formation of isopolymolybdates is significant for recovery of molybdenum from process solutions because both solvent extraction and recduction reactions appear to involve only polymerized molybdenum species. 30 Chariot has stated, for example, that monomeric molybdate ion is reduced infinitely slowly in alkaline solution. Polarographic studies have tended VI . . , . ^ ^ ,.31-33 to confirm this with Mo giving no reduction waves for pH > 5 Similarly the necessity for acidification to pH < 2 for solvent extraction in the USBM process indicates that the extracted species involves isopoly molybdates or even cationic molybdenum species. A review of molybdenum solvent extraction by Zelikman confirms that extraction is most efficient in acid solution and that the extracted species are isopolymolybdates or 34 cations. 2.1.2 Reduced Species and Precipitates For lack of thermodynamic data Pourbaix was able to consider only 3+ Mo, Mo , and M0O2 in his diagram. The predominance area of the mixed-valent molybdenum blue compounds could only be shown approximately based on qualitative observations. It is now well recognized, however, that V IV III Mo , Mo , and Mo species can be prepared and are stable in aqueous 35 solution. Hydrated oxides of each of these valence states can be pre cipitated in appropriate pH intervals. Some progress has also been made in the characterization of the molybdenum blues. VI Like Mo the reduced molybdenum aquo-ions show a tendency towards polymerization. Both Mo11"1" and MoIV can form dimers^ and MoV has been reported to form dimers, tetramers, and higher polymers. The molybdenum blues are undoubtedly polymeric since their formation at a given pH is -In dependent on molybdenum concentration. The reduced molybdenum species are of interest in the present study and the chemistry of each is briefly reviewed in the following sections. Molybdenum Blue Mild reduction of molybdate solutions in the pH range 4 to 0 yields more or less intensely coloured blue solutions. If reduction proceeds beyond the optimum for formation of the blue the colour density decreases 37 and the colour may change from blue to green or brown. There have been many attempts to characterize the blue colloidal precipitates that can be j * i- , 38-42 n n 38 • . prepared from such solutions. Glemser and Lutz and Sacconi and . . 39 Cmi concluded that molybdenum blue was neither a unique compound nor did it represent a definite oxidation state of molybdenum. This conclusion seems to have been generally accepted and is stated in several works on 20,21,43 44 inorganic chemistry. On the other hand Weiser has pointed out that the evidence for the existence of different compounds was based largely on analytical differences of the same order of magnitude as the experimental errors inherent in analyzing a colloidal mass. 41 Arnold and Walker avoided analyzing a colloidal precipitate by extracting the blue into butanol and determining the mean oxidation state of molybdenum by potentiometric titration. They obtained the formula 40 MOg0^7 which agreed with the earlier results of Treadwell and Schaeppi. 42 Ostrowetsky studied the formation of molybdenum blue by mixing V VI solutions .of Mo and Mo in varying ratios and at varying pH's. Total molybdenum in solution was also varied. The formation of the blue was analyzed by spectrophotometry, by isolation of its rubidium salt, and by - 15 -potentiometry. Electrolytic preparation of the blue gave the same results. V VI Optimum conditions for formation of the blue were pH = 1.22, Mo /Mo = 0.5, and [Mo], , = 0.015 M. The blue species was formulated as total Mo, V O Mo. VI 18 The corresponding acid, H Mo 0 , agreed with the results of Arnold and 2. bio Walker and Treadwell and Schaeppi. Decrease of total molybdenum concentra tion inhibited formation of the blue. If a solution of the blue, under the optimum conditions for its formation mentioned above, was titrated with NaOH the solution became 2-green, then brown. The brown solution was found to correspond to HMo.O o 18 and its formation from the blue written as V Mo, Mo, VI °18 2-+ \ MoY 0„4+ + 80H~ 2 4 8 Mo. VI Mo. V°18H 2-+ 2H20 V VI The brown solution could also be prepared directly by mixing Mo and Mo Maximum yield was obtained at pH = 2.7. A third mixed compound, also brown, was found to be formed at V V VI pH = 3 to 4 and Mo /Mo =2.0. It was formulated as Mo. Mo, VI °17H For pH > 4, however, its concentration decreased and at pH = 4.5 a precipi tate of MoO(OH)^ was observed. The work of Ostrowetsky permits a logical interpretation of the colour changes observed on reduction of molybdate solutions in the pH range 0 to 4.5. For pH < 0 the blue compound cannot form and no colour change is - 16 -observed until the Mo stage or lower is attained. For pH between 0 and 4 the first colour to appear is blue for low values of MoV/MoVI. As reduction proceeds the solution becomes green or brown as the two brown MoV/MoVI species form. The solution precipitates MoO(OH) as the MoV stage is approached more closely. If a solution containing any of the mixed valent species is neutralized MoO(OH) is precipitated and molybdates are formed in solution. In addition the mixed valent species can coexist in various propor-VI V tions in solution with varying amounts of excess Mo or Mo depending on the particular conditions. This probably explains the range of mean oxi dation states for molybdenum obtained by different workers using different means of preparation of the colloidal blue. Partial confirmation of Ostrowetsky's work was provided by Filippov 45 and Nuger who observed that the character and intensity of the molybdenum blue spectrum observed during reduction of molybdic acid by hydrazinium chloride varied according to thepH at which the reduction was performed. The maximum absorbence observed by Filippov and Nuger occurred at pH = 1.31 which corresponds to the optimum pH for formation of the blue compound proposed by Ostrowetsky. As the pH of reduction was increased the absorb ence due to formation of the blue decreased. Unfortunately Filippov and Nuger did not record the spectra in the region of absorbence of the two brown species so it is not possible to tell if the absorbences due to these two species increased as would be expected if the overall degree of reduc tion obtained was independent of pH. In the case of Filippov and Nuger's work it is likely that the degree of reduction did depend on pH since for pH > 4 no reduction was observed. The latter ovservation is likely the - 17 -result of a kinetic effect since, as will be shown later, hydrazine in V sufficient excess can reduce molybdate almost quantitatively to Mo for pH between 4.5 and 5.0. Ostrowetsky proposed that since the mixed valent species were hexamers it was likely that they were formed by reduction of hexameric molybdate species. Filippov and Nuger also assumed that molybdenum blues were isostructural with the molybdate species from which they were formed. - 18 - MoV V + + Mo and its chloride salts, R_Mo0Cl,_ (where R = NH., Rb ), can be 2 b 4 46 prepared in acid solution by standard procedures. On neutralization or dilution the green chloride solutions hydrolyze rapidly to give a brown colour for [Mo] > 0.1 M which becomes amber or yellow on increasing dilution. Complete neutralization yields a brown Mo^ precipitate. 47 According to Mellor Klason obtained MoOtOH)^ by adding 3 moles of ammonia to a solution containing one mole of (NH^)^MoOCl^ while Debray found that if excess ammonia was used the precipitate was partially decomposed and the filtered solution contained MoVI. The anhydrous oxide was obtained by several workers by heating the precipitate in vacuo or in inert gas streams. 48 Simon and Souchay performed a detailed spectrophotometric study of V the hydrolysis and concluded that below 2 M HC1 and 3 M H^SO^ Mo was not complexed by the anion of the acid used. Titration of (NH^)^MoOCl^ with NaOH suggested the hydrolysis product immediately before precipitation could be formulated as (HoO^)^ where x indicated an unknown degree of polymeriza tion . 49 Ardon and Pernick confirmed this work concluding that the predomi nant MoV species in dilute HC1, HCIO^, and other acids is a binuclear cation with charge 2+ and is not coordinated to chloride. Their work involved analysis of its elution behaviour from a cation exchange column and cryos-copy of a 0.02 M solution in eutectic HC10„. 4 Subsequently Viossat and Lamache^6 proposed that for [M0V] > 10 2 M the predominant species is a tetramer which transforms into (Mo02+) on dilution. They also reported that if a solution of MoV was neutralized to pH = 1.2 and heated at 80°C a significant fraction of the molybdenum - 19 -precipitated while (Mo02+)2 and (Mo02+)^ coexisted in solution with a previously unknown chestnut coloured species. The chestnut species could be removed from solution on an anion exchange resin. It was found that it was highly polymerized. One manifestation of this polymerization was that small concentrations of it suppressed the polarographic maximum ob served for MoVI in 2 N HCl. Viossat and Lamache proposed that the chestnut V Mo species was actually cationic but had an overall negative charge result ing from strongly adsorbed chloride ions. Only two studies of the precipitation of MoV appear to have been published. They are in fair agreement with each other. Katsobashvili et al. produced MoV in HCl solution using zinc as a reductant. The result ing solutions contained 0.021 M Mo and were 0.03 N in AlCl^. Titration with 0.02 N NaOH showed precipitation was complete between pH 6 and 6.5. Molyb denum remaining in solution was beyond the limit of sensitivity of colori-metric analysis. It was observed, however, that between pH 8 and 10 hydroxyl ions were adsorbed by the precipitate and the molybdenum concentra-VI tion in solution increased. They found that Mo was dissolving from the precipitate so the phenomenon was not simply a question of increasing solu-V VI bility of Mo at a higher pH. Mo also appeared in solution if NH^ was used for neutralization. The amount appearing increased with time and temperature of standing for a given pH and with increasing pH for constant VI time and temperature. It was noted that plots of Mo appearing vs time for constant pH's between 6.5 and 9 had positive intercepts indicating that at least part of the dissolution occurred almost instantaneously on neutral ization. Souchay et al.5"'" neutralized aliquots of 0.03 M MoV solutions in HCl - 20 -with varying amounts of NaOH, agitated them for 10 minutes, and separated the precipitate by filtration. The pH of the filtrate was measured and V Mo remaining in solution was determined polarographically in 6 N HCl after oxidation to MoVI with Ce^+. Sodium and chloride ions in the precipi tate were determined after dissolving it in 3 N HNO^. The results indicated V the precipitation of Mo was complete for pH > 6 but that the consumption of NaOH exceeded the one equivalent expected from the equation (Mo02)4 + 40H~ + 4H20 = 4MoO(OH)3 It was concluded that the precipitate acted as an ion exchanger absorbing Na+ and releasing H+. The pickup of sodium increased when NaCl was added to the solutions. Souchay et al. also observed dissolution of the precipitate but only after a solution of pH = 12.65 was agitated for several hours. It was found by polarographic analysis that after such dissolution the precipitate IV VI contained a mole of Mo for each mole of Mo appearing in solution. The VI V appearance of Mo thus resulted from the dismutation of Mo according to: „ V . IV VI 2Mo • Mo + Mo In 5 N NaOH complete dismutation was almost instantaneous yielding a precipitate in which the mole ratio of molybdenum to sodium was one to one. The reaction proposed was: + 2-2MoO(OH)3 + 30H + Na • MoO + HNaMoO^ + 4H20 which involved formation of an insoluble molybdite identified in a similar IV 51 study of the precipitation of Mo - 21 -IV Mo IV The behaviour of Mo in aqueous solution has been a matter of con troversy. According to Mellor several workers reported obtaining various hydrates of MoC^ by neutralization of reduced molybdate solutions. Con-V versely, Klason maintained these precipitates were probably impure Mo products and hydrated oxides of MoIV did not exist. This opinion was reinforced by Haight and coworkers who concluded, on the basis of polaro-graphic work^2 and analysis of the kinetics of the reduction of MoVI by II 53 IV Sn , that Mo was unstable in aqueous solution and disproportionated III V into Mo and Mo . 54 Guibe and Souchay, however, had previously demonstrated the exist ence of MoIV using polarography. Souchay, Cadiot, and Duhameaux"^ subse-IV quently neutalized a Mo solution to precipitate Mo0(0H)2 and formulated IV + the predominant Mo ion before precipitation as MoO(OH) . IV Ardon and coworkers suggested Mo was actually dimeric based on 56 57 58 ion exchange experiments and cryoscopy. Recently Chalilpoyil and Anson reported that a dimeric assignment is consistent with the behaviour of MoIV during electrochemical oxidation and reduction. Souchay, Cadiot, and Viossat"*^ studied the precipitation of Ho^. IV Varying amounts of NaOH were used to neutralize aliquots of 0.03 M Mo in the same procedure as already described for their work on MoV. Precipi tation began at pH = 1.5 and was complete at pH = 2.8. Sodium and chloride ions could be eliminated from the precipitate by washing with water. The product was analyzed as Mo02 • 2H20. It could be easily dissolved in 2 N IV HC1 to give the spectrum and polarogram characteristic of Mo . After drying at 130°C, however, it took on a grainy appearance and could no longer - 22 -be dissolved in acid. Insolubility in acid is a characteristic of anhydrous 47 Mo02-If neutralization was carried out with an excess of NaOH giving a final pH of 11.5 the product was an insoluble monoalkaline molybdite. The reaction proposed was MoO(OH)2 + MOH = Mo03HM + + + where M = Na , Li , K . In 5 N NaOH the molybdite partially redissolved to IV 2-give Mo in solution. The ion MoO^ was proposed by analogy with molyb date . 59,60 Lagrange and Schwmg obtained an Mo02 • 2H20 precipitate by electrolysis of 0.125 M Na^oO^ solutions on a mercury cathode at -0.6 V (vs SHE) and pH 5 - 6. The precipitate.only formed for pH > 2. At lower pH molybdenum blue was formed and no deposition occurred. The precipitate gave the ASTM diffraction pattern for Mo02 after drying at 500°C in nitro gen. Lamache61 found that electrolysis at -.29 to -.38 V (vs SHE) in an acetic buffer of pH 4.6 gave Mo111 in solution and a precipitate containing V IV equal amounts of Mo and Mo Mo111 Mo111 can be produced in acid solution by reduction with Cd, Zn, Hg or by electrolysis.Mo111 deposits have been formed by electro lytic reduction in alkaline or neutral solution but the rate of deposition 62-66 is slow. No detailed studies of the precipitation of Mo111 to yield Mo(0H)3 - 23 -have been made. Mo(OH).^ and the anhydrous oxide are poorly characterized. 62 Smith verified that electrolysis in alkaline and neutral solutions led to 47 Mo(OH)^ by preparing enough of the deposit for analysis. Mellor reported that Mo(OH)^ dissolves only with difficulty in acids. 64 VI Watt and Davies obtained anhydrous M020^ by reduction of Mo oxide with solutions of potassium in liquid ammonia. They reported that Mo^O^ dissolves readily in 6 N HC1. MO2O2 was converted to insoluble black Mo(OH)^ by agitation in water for 15 minutes at 25°C. Mo(OH) prepared from Mo^O^ was found to have the same properties as Mo(OH)^ pre pared by electrolysis of MoVI solutions at pH 3.1 to 4.4. Neither Mo(OH)^ nor Mo^o^ gave X-ray diffraction patterns. On heating to 325°C in an inert atmosphere, however, Mo(OH)^ gave diffraction patterns for Mo and Mo02 while Mo^Oj still gave no pattern. Messner and Zimmerly16 reduced MoVI solutions with iron at pH 1 to 3.5 and precipitated Mo(0H)3 by neutralization to pH 3.6 to 4.5. The MoO^ produced by roasting the precipitate contained 2.12% Fe and 0.14% Cu. 2.1.3 Summary The preceding review of aqueous molybdenum chemistry suggests that reductive precipitation could lead to various products including mixed valent precipitates. It also illustrates the complex nature of molybdenum species in aqueous solution. Which aqueous species or precipitate predomi nates in a given situation is likely to be dependent on total molybdenum concentration, pH, and potential. Furthermore as these conditions change during a reaction the predominant species and hence reaction path might be - 24 -expected to change. A review of the literature on hydrogen reduction and reduction with hydrazine serves to illustrate these points. - 25 -2.2 Reduction with Hydrogen and Carbon Monoxide Paal and Brunjes^ and Paal and Buttner^ studied reduction of ammonium paramolybdate solutions by hydrogen in the presence of a colloidal palladium catalyst. Figure 5 shows percent reduction to MoIV (calculated on the basis of H^ consumed) vs time for the initial period of reduction carried out at 30°C and atmospheric pressure. The stage was attained after two days at which point the reaction stopped. Reduction was continued by heating to 50° to 60°C and applying a slight overpressure of hydrogen. The reaction proceeded slowly and stopped again after three days when the H2 consumption corresponded to Mo111. The end solution was yellow and the fine black precipitate formed was very difficult to dissolve in cold or hot concentrated HC1 or H2S04. The small quantity that did dissolve gave a light red solution. IV The brown black slurry formed at the Mo stage gave a weight be tween Mo(OH)^ and Mo0(0H)2 when dried in vacuo without heating. When dried with mild heating the weight approximated Mo02. The precipitate was analyzed for molybdenum after dissolution in aqua regia. These results seem to be in accord with those of Souchay et al.^"*" who found that neutralization of MoIV yielded Mo02*nH20 with n very nearly 2. Paal and Buttner did not analyze the precipitate corresponding to the Mo111 stage. The difficult solubility in, and the red colour imparted to, concentrated acid is in accord with the reported properties of Mo(OH)^. The yellow colour of the end solution could have been that of the Mo"^+(H„0). 2 6 58 monomer. The abrupt decrease in the rate of reaction can be attributed to - 26 -Figure 5: Reduction of ammonium paramolybdate solutions by hydrogen in the presence of a colloidal palladium catalyst. - 27 -depolymerization of molybdate occurring when reduction has consumed suffi cient acid. For example the reduction can be written: Mo7C>246~ + 7H2 + 6H+ + 4H20 = 7Mo(OH)4 from which it follows that reduction consumes acid. The paramolybdate itself buffers the reaction in the pH range 5 to 6 by its own dissociation according to: 6- 2- + Mo7024 + 4H20 = 7Mo04 + 8H . It can be shown that after 57% of the total molybdenum has been reduced VI 2-to Mo(OH)4 the remaining Mo exists only as Mo04 . In this particular system the pH is then held between 9 and 10 by the NH4+/NH^ buffer. Since 2-lt seems apparent that MoC>4 is reduced only very slowly the reaction proceeds to completion at a very slow rate. It was found that the precipitates formed absorbed the palladium catalyst almost completely. Lyapina and Zelikmann"*"^ studied hydrogen reduction of 0.05 M sodium molybdate solutions at 100° to 200°C and 10 to 60 atm hydrogen pressure. Figure 6 shows their results for solutions of initial pH 2 and 7 at 40 atm and 200°C. The recommended optimum pH was 2 which corresponds to com-2-plete acidification of MoC>4 . This initial pH assures the reaction is buffered between pH 5 and 6 up to completion. It should be noted, however, that in this work it was found that the solutions contained iron. Thus for lower pH's some of the reduction was performed by the walls of the autoclave. In subsequent runs a quartz liner was used but in these runs various catalysts were also added. The product of reduction was identified - 28 -- 29 -as MoC>2 by X-ray diffraction and analysis for molybdenum. The effect of MoO^ slurry from a previous reduction and metallic molybdenum as catalysts is shown in Figure 7 curve a and Figure 7 curve b, respectively, for inital pH = 3, 40 atm H2, and 200°C. Mo02 was added in a concentration of 67 g/1 and metallic molybdenum as 7% of the amount theoretically necessary for the reaction 2- + Mo + 2Mo04 + 4H = 3Mo02 + 2H20 It was proposed that the reaction of metallic molybdenum with the solution formed "active" Mo02 particles which served as centres of crystallization. It is apparent from Figure 7 that Mo02 from previous reductions was not an efficient catalyst. Figure 7 curve c shows the results for no added catalyst (also, however, with no liner). Complete reduction was obtained in four hours using metallic molybdenum as a catalyst and operating with initial pH = 2, 60 atm H"2, and 200°C (Figure 7 curve d). Under these conditions decreasing the initial molybdenum concentration from 44 g/1 to 5 g/1 decreased the time required for complete reduction from 4 hours to less than % hour. In no case was complete reduction obtained with an initial pH > 2. This is in accord with the assumption that polymolybdates are necessary for an appreciable rate of reduction. 34 In a later review Zelikmann noted for 200°C, 7% of stoichiometri-cally necessary metallic Mo, and 40 g/1 molybdenum that the rate of reduction was linearly dependent on /PH2. He concluded that hydrogen participated in the rate controlling step in atomic form. 12 Sobol studied reduction by both H0 and CO at elevated temperature 30 -ioor (a) initial pH=3, 40 atmH2 ~ slurry catalys t (b) initial pH=3, 40 atm H2 metallic Mo catalyst (c) initial pH=3, 40 otm H2 no catalyst added (d) initial pH = 2 ,6 0 atm H2 metallic Mo catalyst TIME (hour*) 10 Figure 7: Effect of various catalysts, initial pH, and hydrogen pressure on hydrogen reduction of sodium molybdate solutions at 200°C. - 31 -and pressure. Using CO at 80 - 85 atm pressure and 200 - 220°C temperature almost complete reduction of molybdate could be obtained in times approaching one day. This was attributed to the buffering action of formic acid produced according to the reaction CO + E^O = HCOOH. Hydrogen reduction was found to require preliminary acidification to pH = 2 in agreement with Zelikman and Lyapina. It was concluded, however, that the kinetics of gaseous reduc tion were slow and catalysis was required. An MoO^ pulp produced by hydro gen reduction of molybdate solutions was found to have no significant catalytic effect. Sobol, therefore, proposed the use of metallic molyb denum as reductant and his further work did not consider gaseous reduction. Kunda and Rudyk15 proposed hydrogen reduction at 180°C and 23 atm H^ in the presence of 0.025 g/1 PdC^ catalyst for molybdenum recovery from solutions containing about 1 M molybdenum, 1.54 M (NH^)2S°4' and mole ratio free NH^/Mo = 2. Under these conditions essentially complete reduction was obtained in less than one-half hour. The resulting oxide powder absorbed the palladium chloride catalyst but retained enough catalytic activity for use in two "densifications" with fresh molybdate solutions. From the solu tion composition given it is apparent that the hydrogen ion activity in this system was controlled by the NH4+/NH3 buffer. The fact that reduction went rapidly to completion in these conditions can perhaps be explained by the effect of temperature on the molybdate polymerization and NH^/NH^ equi libria. No data is available for the polymolybdates but using free energy 69 data from Barner and Scheuerman's compilation it can be shown that the pH of the NH4+/NH3 buffer falls from 9.27 at 25°C to 5.75 at 200°C. Presumably this fall was sufficient to ensure that the system was buffered in a pH range where polymolybdates could exist. - 32 -The oxide product analyzed 60 - 65% Mo, 4-5% NH3, and 0.5% S compared to 59% Mo for Mo0(0H)3, 75% for Mo02, and 65% for Mo(0H)3- Solid state hydrogen reduction of this product gave 99.9% molybdenum metal. The precise nature of the oxide product was not determined. X-ray diffraction was not mentioned. They did report that when reduction was carried out under milder conditions than those listed above intermediate steps were observed; the solution became marine blue, then molybdenum precipitated as a brown amorphous residue which later agglomerated into black, oval-shaped particles of lower molybdenum oxide. It may be supposed that the brown residue was Mo0(0H)3 and it is possible that in this case IV complete reduction to the Mo state did not proceed. It was observed in the present study that under some conditions Mo0(0H)3 precipitated from molybdenum blue solutions had a blue-black colour rather than its customary brown. 14 Wagenmann patented a process involving hydrogen reduction to recover molybdenum from 1.5 - 3.0 g/1 molybdenum and 60 - 120 g/1 sulfate solutions acidified to pH = 2. Reduction was carried out at 180°C and 20 - 25 atm in a flow through reactor with a residence time of 4% hours. The solution exiting the reduction vessel was cooled to 80°C and expanded to 0.1 atm before filtration to recover the molybdenum product. The filtrate contained 0.05 g/1 Mo thus molybdenum recovery was 97%. The analysis of the product was not given. No catalysis was mentioned. These results are considerably better than those obtained by Zelikmann and Lyapina (Figure 7) at a higher temperature and hydrogen pressure. Since Wagenmann did not describe the construction of his reactor it is not possible to determine whether the walls of the reactor played a role in the reduction. - 33 -2.3 Reduction with Hydrazine The behaviour of hydrazine, N^H^, as a reducing agent has been the subject of much work, the general aim of which was to determine the reasons for the wide variation of stoichiometry with different oxidants and experi mental conditions. Hydrazine was shown to react according to two limiting reactions:* N2H5+ = N2 + 5H+ + 4e~ N H + = JjN. + NH + 2H+ + e~ Z b Z j which could often occur in parallel giving stoichiometries between the two limits. Some oxidizing agents added slowly to boiling highly acid hydrazine solutions also gave a significant yield of hydrazoic acid, HN^. 70 Browne and Shetterly, in summarizing the results of a series of investigations, recognized three classes of oxidizing agents for hydrazine in hot acid solution: class a class b class c produce fairly large little or no HN^ little or no amounts of NH^ and HN3 but much NH^ HN3 or NH3 H2°2 KMnO 4 KI03 KC10„ 4 Mn02 HgO K2S2°8 Fe2°3 Hgci2 *Most of the work to be discussed was performed in solutions of pH < 7 where the protonated form of hydrazine, N„H +, predominates. - 34 -They stated that so many different factors influenced the course of the reactions that it was impossible to establish a simple relationship between the potential of the various oxidizing agents and their ability to produce HN^, NH^, or . 71 Cuy et al. summarized a further series of investigations by sug gesting there were two classes of oxidizing agents. Those undergoing a change of one equivalent per mole reacted according to: N2H5+ • N2H3 + 2H+ + e" followed by one of the two reactions: N2H3 • NH3 + N NH3 + 2N H >• N H 2NH + N2 2 3 4 6 3 If either of these two subsequent reactions occurred faster than further oxidation of N2H"3 a limiting stoichiometry of one equivalent per mole of hydrazine ought to obtain. For oxidizing agents undergoing a change of more than one equivalent per mole a mixed stoichiometry (i.e. between 1 and 4 equivalents per mole of hydrazine) was explained in terms of the genera tion of another oxidizing agent as an intermediate undergoing only one equivalent reduction, for example: 2- + . 4e~ _ Cr„0_ + N-HV > N„ + HCr»Oc + 2H„0 2 1 2 5 2 2 5 2 + 2e~ 3+ + HCr O + 2N H * 2Cr + 2NH. + N_ + 5H„0 2 5 2 5 4 2 2 + -givmg the overall stoichiometry of 1.5 moles K^H,- per mole of Cr^O^ An observed tendency of the stoichiometry, R, (defined as moles of - 35 -electrons/mole of hydrazine) to increase in alkaline solution was thought due to a higher rate of further oxidation of intermediate in alkaline solutions as opposed to a higher rate of its decomposition in acid solutions. 72 Kirk and Browne likewise proposed division of oxidizing agents into two classes: 1. Those that accept only one electron per "active" unit (atom, ion, molecule) were termed monodelectronators; 2. Those that accept more than one electron per "active" unit were termed polydelectronators, i.e. didelectron-ators, tridelectronators, etc. With monodelectronators complete oxidation to nitrogen could be obtained but in cases where R was less than four the sole byproduct was ammonia formed according to: N_H_+ >• N„EL + 2H+ + e~ (slow) Zo 2 3 2NH ——> N .H N„ + 2NH_ (fast) 2 3 4 6 2 3 With didelectronators complete oxidation to nitrogen was predominant but for incomplete oxidation ammonia and a small quantity of hydrazoic acid were formed. The reaction sequence in this case involved initial formation of ^2^2 accordln9 to: NnHc+ • N„H_ + 3H+ + 2e (slow) z b I 2 followed by a series of fast subsequent reactions to produce or HN^ and NH . 3 Oxidizing agents undergoing reduction in more than two stages (complex delectronators) could manifest the characteristics of mono- and didelectronators, for example: - 36 -3+ - 2+ VO + e = VO 3+ - + VO + 2e = VO 73 Higginson et al. proposed that mono- and didelectronators be distinguished by the following criterion: 1. Didelectronators were those reagents which oxidized hydrazine at room temperature in acid solution to produce N2 only. 2. Monodelectronators were those reagents which oxidized hydrazine to NH^"1" and N2 under the same conditions as above with the ratio of NH^"1"/^ depending on relative initial concentrations of hydrazine and oxidant. Simple mechanisms for mono- and didelectronation were proposed: didelectronation N2H5+ 2e~ N2H2 + 3H 2e N2 + 4H monodelectronation N2H5+ le~ N2H3 dimerization 2N2H3 fast 4 6 fast N2 + 2NH3 le N2H2 fast 2e N2 + 4H predominant reaction in acid solution predominant reaction in alkaline solution The stoichiometry for .monodelectronators could lie between R = 1 and R = 4 depending on the relative rates of the two subsequent reactions of N2H3. - 37 -Since in alkaline solutions even monodelectronators produce nitrogen quantitatively the effect of increasing pH was seen as decreasing the rate of dimerization of N^H^ relative to its further oxidation. 74 Subsequently Higginson and Sutton performed an isotopic study 15 involving the oxidation of N enriched hydrazine by excess of various oxidizing agents. The isotopic distribution of the resulting products confirmed the validity of the simple mechanisms proposed. Cahn and 75 Powell independently performed a similar isotopic study with the same results. In a subsequent review Higginson"^ noted that for oxidizing agents in general, yields of ammonia relative to hydrazine consumed are very much smaller in alkaline, neutral, and weakly acid solution than the yields obtained by using monodelectronators in solutions with pH < 3. Thus in alkaline solution few, if any, oxidants give values greater than 0.1 for moles NH_/mole N_H. whereas for pH < 3 most monodelectronators give mole 3 2 4 NH^/mole 1 0.75. Use of the equations: 4 moles NH^ 4 moles + = R moles ^2^4 moles N^H^ 4 moles NH^ moles N2 + moles N_H . moles INLH,, 2 4 2 4 4 (where is defined as due to 4 equivalent reduction) shows that this corresponds to R 2 3.7 in alkaline solution and R £ 1.75 for pH ^ 3. Higginson felt it was likely the change from low to high stoichiometry occurred in the pH range 3 to 5 based on the observation that at pH = 2 - 38 -most monodelectronators gave R < 1.75 while it was observed that at pH = 6 the monodelectronator ferricyanide reacted with R = 4. This prediction was verified for oxidation with manganese trispyrophosphate. For didelectrona tors Higginson proposed the change to high stoichiometry occurred for pH > 0. Browne and Shetterly70 investigated the reaction of aqueous suspen sions of MoO^ with hydrazine. In alkaline solutions the reaction was very slow and some ammonia was produced. In boiling sulfuric acid solutions with excess MoO^ appreciable amounts of ammonia and some HN^ were formed. 72 On the basis of this work Kirk and Browne considered molybdate ions as didelectronators. 77 Jakob and Kozlowski found that molybdate solutions oxidized hydra-VI zine practically completely to - They were able to prepare mixed Mo / MoV compounds by reduction in weakly acid solutions. In slightly more acid V solutions molybdenum blue was formed while at still higher acidity Mo was produced directly without intermediate products. In particular they report-from a solution V VI ed preparation of the Mo /Mo compound NH V VI Mo^O Mo O^ (OH), containing 0.317 M molybdenum as ammonium paramolybdate and 0.06 M N2H4* This compound has the same MoV/MoVI ratio as the brown species reported by 42 Ostrowetsky between pH 3 and 4. In fact if the formula given by Jakob and Kozlowski is rewritten to contain six atoms of molybdenum and water is removed it becomes Mo, Mo, V 17 and, if it is assumed to be protonated in solution, - 39 -V Mo . 4 OH which is the composition found by Ostrowetsky. 31 78 Holtje and Geyer and Rao and Suryanarayama reported that under V no conditions does hydrazine reduce molybdate solutions beyond the Mo state. It was thus recommended as a method for the preparation of MoV VI stock solutions. The latter authors reduced 0.4220 moles of Mo in 1 N HC1 on the boiling water bath with 0.154 moles of hydrazine. They were interested only in obtaining quantitative yield of MoV so did not determine the excess hydrazine. These results, however, indicate they obtained R - 2.74. Since, according to Higginson, a monodelectronator should VI approach R = 1 under such conditions these results suggest Mo behaved as a didelectronator. 45 Fillipov and Nuger reported that at pH > 4, 0.008 M molybdate solutions were not reduced by 0.02 to 0.04 M hydrazine. At pH 1.3 to 1.4 only molybdenum blue was formed. 79 Ostrowetsky and Brinon studied reduction of molybdate solutions by hydrazine in 0.1 to 10 N acid solutions. In this range of acidity mixed MoV/MoV]" compounds did not form and the only product was MoV. The rate of of the reaction increased with increasing temperature and decreased with increasing acidity. Detailed investigations were made in 2 N HC1 at 0°C where the reaction was slow enough to follow without difficulty. Spot tests showed that the solutions contained hydroxylamine and polarography in 4 N NaOH allowed determination of the total concentration of ^H^ + + VI V NH^OH . Mo and Mo were determined by polarography in 2 N HC1. - 40 -It was found that the total concentration of N„ H„ + NH OH: did .not change 2 4 3 V during the reaction and that a plot of Mo vs initial mole ratio [N^H^]/ [Mo] showed a break at mole ratio 0.5. The reaction proposed was + + 2e~ , V „ + N-H.. + HMo_0. (Mo 0o ) „ + 2NH OH Z o Z b Z Z 3 + 80 The species HMo2Og had been proposed by Chauveau et al. as the predomi-VI nant Mo species in 2 N HCl. In the same conditions of temperature and acidity, however, it was found that for the stoichiometric ratio (according to the above equation) of [N2H<_ + ] / [MoVI] = 0.5 an increase of initial concentration of MoVI from 0.02 M to 0.24 M changed the reaction so that N2 as well as other unidenti fied products resulted. The reactions proposed in this case were: 2HMo_0 * + N_H_+ + H+ = Moyl004+ + N_ + 3H„0 Z b Z i> 4 B Z Z or H Mo 0 _ + N H + + 3H+ = Mo 0Q4+ + N + 5H 0 Z 4 13 Z o 4 o 2 Z It was reasoned that the change in the reaction was due to a change in the degree of polymerization of either MoVI or MoV. Chauveau et al.^° had reported an equilibrium between HMo„0 * and H Mo.O., _ while the work of 2 6 2 4 13 Viossat and Lamache"^ indicated that tetrameric MoV could exist in equilibrium with a dimeric form. At a given acidity, therefore, an increase in concentra tion of molybdenum would favour the tetracondensed forms of the MoV"C and MoV species and hence a four equivalent oxidation of hydrazine. Ostrowetsky and Brinon's work appears to be the only report of 81 oxidation of hydrazine to hydroxylamine. Audrieth and Ogg in their review of oxidation of hydrazine stated that no evidence had been reported to demon strate that hydroxylamine is an oxidation product of hydrazine. - 41 -82 VI Huang and Spence investigated the reaction of hydrazine and Mo at 70°C in phosphate buffers of pH 1.2 to 3.2. The initial mole ratio of + VI VI -4 N^Hj. to Mo was maintained at 0.5 and Mo was varied between 5x10 M -4 and 7 x 10 M. The hydrazine was oxidized quantitatively to and the reaction was first order in each reactant. The rate of reaction increased with increasing pH and exhibited an order in H+ of 0.25. It was suggested that the fractional dependence on H+ could be due to changes in polymeriza-VI VI tion of Mo with pH, ionization of a monomeric Mo species, or involve ment of H+ in a rate controlling step. N2H2 waS ^etectec^ qualitatively by mass spectrometry and trapping with unsaturated acids. The observed stiochiometry and the presence of ^H,, indicated MoVI behaved as a didelec-tronator. 8 3 Nusgra and Sinha observed that in 1 N H^SO^ 0.025 M hydrazine and 0.01 M MoVI reacted to yield N quantitatively. They detected both N2H3 and ^2^2 depending on tne relative amount of MoVI added to hydrazine solu tions . The preceding summary indicates that MoVI tends to oxidize hydrazine to nitrogen with R - 4. This has been interpreted in terms of predominant didelectronation to yield N2H2 as a first step and also by series of mono-delectronation steps. Near-quantitative oxidation to N2 by one electron steps, however, appears to be unlikely except in alkaline solution. Two electron steps require that the molybdate species involved be reduced to MoIV if they are monomeric. Mechanisms based on the assumed chemistry of MoIV must, however, be open to question until more is known about the chemistry of MoIV. VI Since Mo tends to polymerize it seems reasonable to assume that the nature of the reaction could depend on the particular MoVI species involved. - 42 -As has been seen this concept was used by Ostrowetsky and Brinon to explain a change in the nature of the reaction with increasing molybdenum concentra tion in acid solution. There have been no detailed studies of the reaction under condition of concentration and pH where molybdate polyanions predomi nate in solution and where formation of mixed MoV/Mo^^~ species might be expected to occur. The work of Jakob and Kozlowski does, however, suggest that four electron stoichiometry and at least intermediate formation of V . VI Mo /Mo species are to be expected. - 43 -2.4 Reduction with SO and H S 31 Holtje and Geyer reported that SC^ reduces weakly acid molybdate solutions to give light green or blue solutions in which only 0.5% of the V molybdenum is present as Mo . In more concentrated acid no noticeable 88 reaction occurred. Wardlaw made similar observations for both SC^ and I^S. Thus at best it appears that neither of these reductants can take molybdate past the mixed valent molybdenum blue stage. - 44 -2.5 Summary The work reported in the literature indicates that reductive precipi tation of molybdenum oxides with hydrogen or hydrazine might be technically feasible. Hydrogen and hydrazine, as a four electron reductant, are clean reductants thus they offer the potential of producing a pure oxide product. Most of the work reported, however, has been performed in acid solutions and the action of these reductants on neutral or alkaline molybdate solutions is not well documented. In the neutral pH range it is apparent that polymerization of molyb date ions might be a factor in any reduction reaction. In addition mixed-VI V valent Mo /Mo compounds might be expected to form in this range and these could influence the course of the reduction reaction. In the case of reduc tion with hydrazine the reaction might be able to follow different paths depending on pH and degree of polymerization of molybdate. Ill IV V It has been seen that Mo , Mo , and Mo hydrated oxides as well as mixed valent oxides can be precipitated in neutral solution. There is also a possibility that the MoV precipitate may act as an ion exchanger. - 45 -3. SCOPE OF PRESENT WORK The present work was undertaken to investigate the use of hydrogen and hydrazine for reductive precipitation of molybdenum oxides from solutions produced by a sodium hypochlorite leach of Cu-Mo rougher concentrates. The specific objectives were to define the pH range in which reduction could be achieved, the kinetics of the reduction reactions, and the nature of the oxide products. The procedure adopted was first to determine the behaviour of the reactions using sodium molybdate solutions then to investigate the effects of other components of the actual leach solutions, in particular copper and sodium chloride. - 46 -4. EXPERIMENTAL Molybdate solutions for reduction by both hydrogen and hydrazine were made up with Mallinckrodt analytical reagent sodium molybdate (Na2MoO^"2H2O) . The manufacturer's assay was 99.5% sodium molybdate minimum and the chemical was used without further purification. The molybdate solutions were standardized by flame atomic absorption spectro photometry and gravimetric analysis for molybdenum. Mo^ stock solutions for polarographic calibration curves and neutral ization experiments were produced by reducing 3.5 N HC1 sodium molybdate solutions by shaking in a flask with metallic mercury. The solutions were 46 i standardized by titration with eerie sulfate using ferroin as an indicator. ' Hydrazine solutions were made up by dilution of BDH 99-100% hydrazine "hydrate or Eastman Kodak 64% hydrazine in distilled water. The hydrazine 85 solutions were standardized by potassium iodate titration in 5 N HC1. Commercial tank hydrogen was used for hydrogen reduction experiments. Hydrogen reduction was performed in a Parr 2 litre autoclave. A glass liner was used for all runs. All parts of the bomb contacting the solution were titanium. The resistance heater surrounding the bomb was controlled by a Yellowsprings Instrument Company Thermistemp Temperature Controller Model 71 and a Variac. The Variac was set to give the approxi mate temperature desired and the Thermistemp maintained the desired tempera ture by operating a relay which reduced the power input to the heater by about 15% in the cooling cycles. It was found that on pressurizing the bomb with hydrogen there was a significant temperature increase (e.g. pressurizing to 30 atm H at 180°C led to a temperature increase of around 30°C). To - 47 -reduce this effect additional cooling was provided by a jet of compressed air or nitrogen introduced at the base of the bomb. The gas flow was con trolled by a solenoid valve which opened on the cooling cycle of the Thermistemp. In all runs stirring was maintained at 600 rpm. In a typical run one litre of sodium molybdate solution adjusted to the desired pH was added to the glass liner and the bomb was sealed. Nitrogen was bubbled through the solution for about five minutes and the bomb was pressurized to 50 psig with nitrogen. The bomb was brought rapidly to just below the desired temperature using full power input to the furnace. The Variac setting was then reduced and the temperature of the run controlled by the Thermistemp. Once the temperature was stable hydrogen was admitted to the bomb. The hydrogen pressure was increased to that desired and main tained throughout the run. The variation in temperature during a run was ±5°C and pressure ±5 psig. Samples were taken at appropriate times throughout each run. The sampling system was cleared before each sample by discharging at least 20 ml of solution. Agitation was maintained during sampling. Samples were cooled either by standing at room temperature or holding under cold tap water. After each run the apparatus was cleaned by running for one hour at 100°C with 50% nitric acid in the glass liner. Hydrazine experiments were performed in a 420 ml pyrex vessel. The reactor was maintained at the desired temperature by immersion in a water bath. Temperature control was ±0.3°C. In a typical run 200 ml of sodium molybdate solution adjusted to the desired pH with HCl was brought to temperature in the reactor. An equal - 48 -volume of hydrazine solution adjusted to the same pH was preheated in a separate flask. At zero time the hydrazine solution was added to the molyb date solution. Stirring was maintained at 300 rpm for each run. The pH was controlled by additions of 1:1 HC1 as required during the run. Samples were taken by pipetting 15 to 30 ml of the reacting slurry from the reactor. Various methods were used to freeze the reaction to permit determination of the concentrations of molybdenum and hydrazine as functions of time. Since preliminary runs had shown that the reduction reaction did not occur above pH 6.5 the first method adopted was to discharge the sample into a centrifuge tube containing enough concentrated NaOH to raise the pH to V between 8 and 12. It was hoped this would also precipitate all of the Mo formed. The second method used was to discharge the sample into a porcelain filter crucible and collect the filtrate in a centrifuge tube again con taining enough NaOH to raise the pH to 8 to 12. This method of analysis gave significantly different results. The third procedure was to discharge the sample into a small beaker and quickly titrate it with 0.1 to 1 N NaOH to obtain a particular pH. The resulting slurry was then filtered and the filtrate analyzed for molyb denum and hydrazine. A fourth procedure was simply to discharge the sample into a beaker containing an equal volume of water at about 0°C. Preliminary runs had shown the reaction was very slow at room temperature so it was felt this procedure would permit time for subsequent filtration. The fifth technique used was to discharge the sample into a filter - 49 -crucible then dilute the filtrate to a concentration suitable for determina tion of molybdenum by atomic absorption (-10 ug/ml). A few runs were performed to follow the evolution of nitrogen gas during reaction. The reaction was carried out in a sealed flask and the gas evolved collected over mercury. These runs were carried out at room tempera ture. To begin a run a hydrazine solution adjusted to pH = 4.5 was added to a molybdate solution of the same pH in the flask. The flask was quickly stoppered and stirring begun. The pH was not controlled during these runs. Molybdenum in solution was determined both by flame atomic absorption spectrophotometry using a Perkin Elmer Model 306 spectrophotometer and by polarography using a Sargent Model XXI polarograph. Solutions for polar-ography were diluted to contain between 0.05 and 0.15 g/1 Mo while those for atomic absorption were diluted to contain between 0.01 and 0.04 g/1 Mo. Dilution for atomic absorption was performed with a 10% AlCl^, 5% NH^Cl 86 solution which was found by Ismay to eliminate interferences. Polarography was performed in 2 N HCl where the waves of MoV and MoVI are distinct.79 Atomic absorption gave total molybdenum regardless of the proportions V VI of Mo and Mo in the sample and hydrazine did not interfere. Polarography VI V of Mo /Mo mixtures in the presence of hydrazine was unsuitable for separate V VI determxnation of Mo amd Mo because at the acidity requried (2 N HCl) the reduction of MoVI by hydrazine proceeded at a rate sufficient to make V the determination of Mo of doubtful accuracy. Total molybdenum could be determined. Atomic absorption and polarography agreed within 2%. Hydrazine in sample filtrates was determined by titration with 0.1 N KIO-j in 5 N HCl. The end point was marked by the disappearance of the pink iodine colour from a CCl^ layer in the titration flask.^5 Ammonia, - 50 -hydroxylamine, and Mo did not interfere. When appreciable hydroxylamine was present in the titration solution, however, a faint pink colour returned V to the CCl^ layer on standing for a few hours. Mo did interfere in the titration but in general the amount of MoV present in the reaction filtrates was small enough to render a correction unnecessary. Ammonia in sample filtrates was determined by a method described by 8 7 DeVries and Gantz. Hydrazine was first oxidized by 0.4 N KIO^ as pre viously described. The I+ and excess 10^ were reduced to I by excess Na^SO^. The excess sulfite was then removed by bubbling air for 15 minutes. Ammonia was then determined by distillation into 0.1 N HC1 after making the solution basic. Hydroxylamine did not interfere. Precipitates produced by hydrazine reduction were analyzed for total 89 molybdenum gravimetrically by precipitation of molybdenum as PbMoO^. The mean oxidation state of molybdenum was determined by oxidation with excess eerie sulfate and back titration with ferrous ammonium sulfate using ferroin 46 as an indicator. Water was determined using a Dupont 950 thermogravimetric analyzer. Chlorine was determined by dissolving a sample of the precipitate in 3 M HNO^, addition of excess AgNO^, and back titration with HCl using a silver electrode and a potassium sulfate reference electrode."^ Sodium was determined by flame emission photometry. Precipitates resulting from hydrogen reduction were characterized by X-ray diffraction and thermogravimetric analysis. - 51 -5. RESULTS  5.1 Hydrogen Reduction Hydrogen reduction experiments were performed at 200° to 220°C with 30 atm pressure of . The initial concentration of molybdenum and the initial pH were 17 g/1 and 2, respectively. For times up to 7 hours reduction proceeded only to the molybdenum blue stage and no precipitation occurred. An experiment was then performed in the presence of a platinum clad expanded niobium mesh. In this case the solution passed through the molyb denum blue stage in 30 minutes and later samples were only slightly coloured. Figure 8 shows a plot of percent reduction vs time. The final pH was 8. The black precipitate formed only slightly plated the apparatus and was easily filterable. It gave an X-ray diffraction pattern identical to ASTM 5-0452 for Mo02• Thermogravimetric analysis showed the precipitate was anhydrous. Experiments with the mesh were not continued because the niobium substrate suffered extreme hydrogen damage and broke under its own weight. The run with the mesh resulted in some precipitation on the stirrer and other fixtures as well as the glass liner. A further run performed without cleaning the apparatus resulted in formation of a blue precipitate giving an X-ray powder diffraction pattern indicating the presence of Mo02 90 and MoO^ (ASTM 21-569) . Ceric sulfate titration indicated the average oxidation state of molybdenum in this precipitate was 5.53. Only a small amount of precipitate was formed after 12 hours. The dark blue solution contained 16 g/1 molybdenum with an average oxidation state of 5.67. The apparatus was then cleaned to remove all deposits and the run was - 52 -20 L_ • L 0 5 10 TIME (hours) Figure 8: Reduction of 17 g/1 sodium molybdate solution by hydrogen at 200°C and 30 atm H2 in the presence of Pt clad niobium mesh. - 53 -repeated. Almost no reduction occurred. The solution was coloured light blue and there was no precipitate. Three runs were performed in 3 M NaCl to simulate solutions from the proposed sodium hypochlorite leach. Minimal reduction was obtained even when using metallic molybdenum powder as a catalyst. Hydrogen reduction experiments were discontinued when it was observed that the liquid condensing between the glass liner and the bomb severely corroded the titanium bomb. - 54 -5.2 Reduction with Hydrazine Preliminary runs were carried out at 50°C using 5 g/1 molybdenum solutions and various hydrazine concentrations. For pH ~ 5 reaction occurred almost immediately on mixing the reactants. The solution under went a series of colour changes. It first became blue, then green, and finally an opaque brown. A brown precipitate formed during the green stage but not during the initial blue one. The effect of pH was investigated at 50°C, 5 g/1 initial molybdenum concentration, and an initial hydrazine to molybdenum mole ratio of 4:1. For pH > 6.5 no reaction occurred. As the pH was decreased below 6.5 the rate of precipitation increased but for pH < 3 the nature of the reaction changed. The initial blue colour remained throughout the reaction and the rate of precipitation decreased. The precipitate formed was blue rather than brown. More detailed investigations were performed at pH = 4.5, 50°C and 4:1 initial mole ratio of hydrazine to molybdenum. These conditions gave a convenient reaction rate for analysis. The results obtained, however, were found to depend markely on the sampling method employed. Figure 9 curve a is a plot of the concentration of molybdenum remain ing in solution vs time obtained by neutralizing 20 ml samples of the re acting slurry with 8 N NaOH and then centrifuging to obtain a supernatant solution for analysis. Curves b through e were obtained by neutralizing 20 ml samples of the slurry by titration to the pH's indicated followed by filtration through porcelain filter crucibles. Curve f was obtained by discharging the sample of slurry into 20 ml of cold water followed by - 55 -I I I I L_ 0 30 60 90 120 TIME (minutes) Figure 9: Effect of sampling technique on concentration of molybdenum remaining in solution vs time at 50°C with pH = 4.5, initial molybdenum concentration =5 g/1, and initial hydrazine to molybdenum mole ratio =4:1. " - 56 -filtration. Figure 10 compares the results obtained using methods involving direct filtration of the slurry samples with those obtained by first dis charging the slurry into an equal volume of cold water. The open circles represent the latter technique. The triangles represent points obtained by direct filtration of the slurry and dilution of an aliquot of the filtrate for atomic absorption analysis of molybdenum. The crosses represent points obtained by direct filtration of the slurry samples into centrifuge tubes containing 1 ml of 8 N NaOH. All three of these methods gave similar results for molybdenum remaining in solution. When sampling was performed by filtration into centrifuge tubes con taining 8 N NaOH it was observed that the clear light brown filtrate formed a brown or blue-green precipitate on mixing with the caustic. Figure 11 shows the dry weight of precipitate retained on filtering 25 ml samples of slurry through a porcelain filter crucible, the dry weight of the precipi tate formed in the centrifuge tube, and the concentration of molybdenum remaining in solution in the centrifuged filtrate. Figure 12 shows the results obtained under the same conditions (50°C, pH = 4.5, initial molyb denum concentration = 5 g/1) but with a 2:1 initial mole ratio of hydrazine to molybdenum instead of 4:1. Since the above sampling method was rather tedious it was decided to use the simpler procedure of direct analysis of the sample filtrate for molybdenum remaining in solution in order to investigate the effects of temperature and initial hydrazine concentration on the rate of the precipi tation reaction. . The initial molybdenum concentration was ~5 g/1 and the pH = 4.5 for each run. Figures 13 and 14 show the results obtained. - 57 -pH = 4.5 T=50°C inilibl Mo = 5 g/1 initial N2H4= 6.6g/l o discharge into cold water then filtration a direct filtration into caustic A direct filtration then dilution 20 TIME (minutes) 30 Figure 10: Comparison of results obtained by direct filtration of slurry samples with those obtained by previous dilution of samples with an equal volume of cold wa'ter. - 58 -TIME (minuteS' ) Figure 11: Distribution of molybdenum between solution and precipitates as a function of time at 50° C with pH=4.5, initial molybdenum concentra tion =5 g/1, and initial hydrazine to molybdenum mole ratio =4:1. * - 59 -0 20 40 60 TIME (minutes) Figure 12: Distribution of molybdenum between solution and precipitates as a function of time at 50°C with pH = 4.5, initial molyb denum concentration =5 g/1, and initial mole ratio of hydrazine to molybdenum - 2:1. - 60 -- 61 -0 10 20 30 TIME (minutes) Figure 14: Effect of initial mole ratio of hydrazine to molybdenum on rate of precipitation at 50°C with pH = 4.5 and initial concentration of molybdenum =5 g/1. - 62 -In order to simulate the conditions expected in treating solutions from the sodium hypochlorite leach a run was carried out in 3 M NaCl. Figure 15 compares the results of this run with one performed under the same conditions but without NaCl addition. The effect of copper on the reaction is shown in Figure 16. The presence of copper did not affect the rate of precipitation of molybdenum. Some of the copper was also precipitated. The precipitate itself changed from the normal brown colour to deep blue on drying for a few minutes in air. It can be seen from Figures 14 and 15 that for low initial mole ratios of hydrazine to molybdenum and in the presence of 3 M NaCl there was a significant time before precipitation began. In each case, however, reaction began immediately on mixing the reactants as evidenced by the appearance of the colours characteristic of the mixed valent molybdenum blue species. In the extreme case of a 1:1 initial mole ratio of reactants no precipitate was formed in 1 hour at 50°C and pH 4.5 even though some reduction had occurred. The apparent reaction order in molybdenum was 1.5 as shown in Figure 17 which was plotted using the data of Figure 14. The slopes of the lines in Figure 17 were used to determine the order in ^H^. Figure 18 is a plot of In slope vs In [N2H ]• From Figure 18 the order in N2H4 1 5 was 1.64. Writing the precipitation reaction as -d[Mo] = k[Mo] 1.64 [N^H^] " allowed determination of the rate constant k from the slopes of Figure 17. The values of k so obtained are given in Table III. The value obtained for initial hydrazine concentration = 0.522 mole/1 5.oh 60---- 63 -4.0 3 M No Cl 3.0 < or Z LU O o u o 2 2 .0 no NoCI 1.0 -L 3 4 TIME (hours) Figure 15: Effect of 3 M NaCl on rate of precipitation at 50°C for pH = 4.5, initial molybdenum concentra tion =5 g/1, and initial mole ratio of hydrazine to molybdenum - 4:1. - 64 -1 i i : i_ 0 2 0 4 0 60 TIME (minutes) Figure 16: Effect of addition of 0.55 g/1 Cu as copper sulfate on the rate of precipitation of molybdenum at 50°C for pH = 4.5, initial molybde num concentration =5 g/1," and initial hydrazine to molybdenum mole ratio - 4:1. - 65 -0 10 20 30 TIM E (minutes) Figure 17: 1.5 order in molybdenum plots for T = 50°C, pH = 4.5, and initial molybdenum concentration =5 g/1. - 66 -—1 1 I I I I L_ -1.8 -1.4 -1.0 -0.6 LN INITIAL HYDRAZINE CONCENTRATION (mole/1) Figure 18: Order in N H based on 1.5 order in molybdenum. - 67 -initial hydrazine to molybdenum mole ratio , -2.14 2.14 . -1 k mole 1 mm 3:1 3.35 4:1 3.55 5.5:1 3.32 7:1 3.07 Table III Rate Constants Obtained for Different Values of Initial Hydrazine to Molybdenum Mole Ratio (10:1 initial mole ratio of hydrazine to molybdenum) has not been included because the reaction was more rapid than could be followed accurately with the sampling procedure used. The temperature dependence data of Figure 13 is plotted in Figure 19 assuming 1.5 order in molybdenum concentration. In each case pH = 4.5, initial molybdenum concentration =5 g/1, and initial mole ratio of hydrazine to molybdenum = 4:1. Figure 20 is an Arrhenius plot based on the slopes of the lines from Figure 19. The activation energy obtained is 14.1 kcal/mole. Gravimetric analysis of the brown precipitate after drying overnight in an evacuated dessicator yielded 57.5% and 56.5% molybdenum (expected for MoO(OH)^58.9%). Ceric sulfate titrations assuming all molybdenum was V present as Mo yielded 54.0% and 52.0% molybdenum. The thermogravimetric weight loss curve obtained in a helium atmos phere is shown in Figure 21. The observed weight loss was 15.8%, compared to 16.57% expected for loss of water by MoO(OH) . - 68 -0 20 40 60 80 100 TIME (minutes) Figure 19: Effect of temperature on rate assuming 1.5 order in molybdenum for pH = 4.5, initial molybdenum concentration =4 g/1, and initial mole ratio of hydrazine to molybdenum =4:1. I I I I J I I L_ 1—— 0 200 400 600 8 00 TEMPERATURE (°C) Figure 21: Thermogravimetric weight loss curve for brown precipitate produced by reduction with hydrazine. - 71 -Precipitate formed at pH = 4.5 and 50°C was found to contain 0.42 ± 0.09 percent sodium. Neutralization of the reactor slurry to pH = 7 with NaOH after completion of a run at pH = 4.5 and 50°C gave a precipi tate containing 0.61 ± 0.03 percent sodium while neutralization to pH = 8 resulted in 1 percent sodium content. Precipitate formed at 50°C and pH = 4.5 in 3 M NaCl solution contained 3.3 percent sodium. No chloride ion was found in any of the precipitates. Figure 22 shows the concentrations of molybdenum and hydrazine remaining in solution vs time at 50°C with pH = 4.5 and initial mole ratio of hydrazine to molybdenum of 4:1. Although the hydrazine concen tration remained constant after about one hour the molybdenum concentration continued to fall. After standing overnight at room temperature the molyb denum concentration in solution reached 0.053 g/1. This compares to 0.041 g/1 obtained by neutralizing a stock MoV solution to pH = 4.5 at 50°C. Similar behaviour was observed in other runs. This behaviour introduced some uncertainty into the computation of the stoichiometry of the reaction. The method finally adopted was simply to use the hydrazine concentration remaining at the end of each run when the molybdenum concentration was reduced to less than 0.4 g/1. Complete reduction of Mo^ to Mo^ was assumed even though, as will be discussed, VI V the molybdenum remaining in solution was present as mixed valent Mo /Mo species. Since very little molybdenum actually remained in solution this procedure did not make much difference to the stoichiometry obtained. The average stoichiometry based on nine runs was 1.55 moles molybdenum reduced per mole hydrazine consumed. The extreme values were 1.47 and 1.77. This stoichiometry suggested that molybdenum was acting as a I i I I : J 1 0 1 2 3 4 5 TIME (hours) Concentrations of hydrazine and molybdenum remaining in solution vs time at 50°C with pH = 4.5, initial molybdenum concentration =5 g/1, and initial hydrazine to molybdenum mole ratio =4:1. Figure 22: - 73 -monodelectronator. Based on the literature the expected reaction products from monodelectronation of hydrazine are nitrogen and ammonia. To verify the stoichiometry a run was performed in which the evolved gas was collected and the filtrate analyzed for hydrazine, molybdenum, and ammonia. The -3 volume of solution used was 200 ml atm 6.95 x 10 moles N2H4 were consumed -3 generating 1.616 x 10 moles of nitrogen. Only a trace of ammonia was found and the pink colour did not return to the CCl^ layer in the hydrazine titration solution on standing. Gas chromatography of the collected gas did not detect hydrogen. In every run the end filtrate was coloured indicating the presence of mixed valent MoV/MoVI species. This was verified by polarography in 2 N HCl and in 5 N NaOH. In 2 N HCl the waves of MoV and MoV]" were distinct but quantitative determination of the ratio of MoV to MoVI was not possible because in 2 N HCl the excess hydrazine present reduced the remaining MoVI at a significant rate. Polarography in 2 N HCl, therefore, served only to VI confirm that Mo was indeed present in coloured filtrates. Polarography in 5 N NaOH made use of the observation by Souchay et al.5"'" that MoV dismutes into MoVI and Mo"I'V in such solutions. Since MoVI does not give any polarographic wave in basic solution the observation of a reduction wave indicated the presence of MoIV resulting from the dismutation. This technique showed that the MoV observed on polarography in 2 N HCl was not due simply to the reduction of MoVI by hydrazine in 2 N HCl. Polarography in 5 N NaOH did not prove suitable for quantitative determination of MoV. The height of the MoIV~ wave observed was sensitive to time before polar ography and the solutions tended to drop out a grey-green precipitate on standing. - 74 -6. DISCUSSION The hydrogen reduction experiments confirmed the importance of heterogeneous catalysis and pH in attainment of reasonable rate of reduc tion. The only run in which significant reduction was obtained was that performed in the presence of the platinum clad niobium mesh catalyst with a solution acidified to pH = 2. The rate of reduction was similar to that obtained by Lyapina and Zelikman"*"^ using metallic molybdenum as a catalyst (Figure 7 curve d). Even in this run, however, complete precipitation of molybdenum was not obtained. The failure to obtain complete reduction was probably due to the increase of pH to 8 and depolymerization of the molyb denum remaining in solution. This depolymerization would result in a 68 decrease of the rate of reduction like that found by Paal and Buttner and shown in Figure 5. The results of the present study and those reported in the literature cast some doubt on Wagenmann's claims. It seems probable the walls of his apparatus were involved in the reaction. Satisfactory performance of hydrogen reduction would thus require sufficient acidification to maintain polymerization of MoVI species. In a batch process this could be accomplished by initial addition of acid beyond the H^MoO^ point or by pH control during the reduction. In addition a catalyst and an autoclave material resistant to the conditions encountered in hydrogen reduction of high temperature chloride solutions would be required. The experiments using hydrazine as a reductant again reflect the importance of polymeric molybdenum species, and hence pH and total molybde num concentration, in the reduction and precipitation reactions. Above - 75 -pH =6.5 monomeric MoVI predominates and no reduction is observed. On decreasing pH reduction occurs and the rate of precipitation reaches a maximum at about pH = 4.5. This can be attributed to formation of reduc ible polymolybdates. Below pH -4.5 the rate of precipitation decreases because in this pH range intermediate mixed valent Mo^/Mo^ compounds are stable. At pH = 4.5 the sequence of colour change observed during reduction 37 is consistent with that described in the literature and interpreted by 42 Ostrowetsky. Thus on reduction of polymolybdates the initial reaction products are coloured mixed valent MoVI/MoV species. At pH = 4.5, however, these species are not stable and may be reactive towards further reduction to yield a MoV precipitate or tend to hydrolyse yielding a MoV precipitate and MoVI in solution. At pH = 3 where mixed MoV/MoVI species are more stable the rate of precipitation decreases. The curves of Figures 11 and 12 illustrate qualitatively the import ance of mixed valent species in the precipitation reaction. The curves showing the weight of precipitate formed in the centrifuge tubes represent MoV present in mixed valent species. That is, when the sample filtrate mixed with the caustic in the centrifuge tubes the mixed valent species were made unstable by the increase of pH and decomposed to yield MoV precip-VI , itate and Mo in solution. (The pH in fact was likely high enough to cause some disproportionation of MoV to give MoIV in the precipitate and VI additional Mo in solution. This phenomenon rendered the results of qualitative value only.) It can be seen in Figures 11 and 12 that the maximum rate of precipi tation approximately coincides with the maximum concentration of mixed - 76 -valent species. This suggests that precipitation occurs through further reduction of these species to yield a MoV precipitate. Polarography indeed confirmed that the slightly coloured solutions at the end of each V VI VI run contained both Mo and Mo . Complete reduction of Mo is never obtained because as the total molybdenum in solution reaches a low enough value the equilibrium between polymeric and monomeric MoVI ensures that 2-a residual concentration of irreducible MoO^ always exists. The minimum residual concentration of molybdenum remaining in solution obtained in -4 this work was 0.053 g/1 (5.5 x 10 M). This is of the same order as the 2-concentration of MoO^ in equilibrium with polymeric molybdate species at pH = 4.5 (Figure 4). The formation of mixed valent species also explains the observation of a significant time lag between beginning of reduction and beginning of precipitation for low initial mole ratios of hydrazine to molybdenum and at low temperature and in the presence of 3 M NaCl.(Figures 13, 14, 15). In the first two cases no precipitation occurs until reduction has proceeded far enough to yield an appreciable concentration of the mixed valent ion that is ultimately reduced to yield MoV precipitate. Lower mole ratios of reactants and lower temperatures delay the attainment of this concentration. The effect of NaCl may be due either to stabilization of mixed valent species of lower degrees of reduction than the species that is reducible V to Mo , or to stabilization of this ion itself. V VI The properties of mixed valent Mo /Mo species may also be the reason for the behaviour shown in Figure 22 where reduction, as represented by the curve of hydrazine concentration vs time, stops while precipitation V of Mo continues. Presumably this is due to the slow decomposition of - 77 -VI V V VI mixed valent Mo /Mo species to give Mo precipitate and Mo in solution. As has been mentioned, for the low total concentration of molybdenum VI present, a significant fraction of this Mo would be present as irreducible 2-monomeric MoO^ so consumption of hydrazine through further reduction would be negligible. The effect of sampling procedure on the curves of molybdenum vs time as shown in Figures 9 and 10 can be interpreted in terms of the ability of V IV VI Mo to disproportionate into insoluble Mo and soluble Mo . The curves of Figure 9 were generated by neutralizing slurry samples before filtration. As can be seen increasing the pH of neutralization increased the amount of molybdenum found in the filtrate for any given time. Obviously dispro-portionation occurred at the instant of neutralization as observed by 50 Katsobashvili and the higher the pH the greater the extent of dispropor-tionation. It is clear that the precipitate itself was disproportionating because filtration before neutralization gave results indistinguishable from those obtained with no neutralization followed by filtration (Figure 10) . It is apparent that the kinetic results obtained can be qualitatively VI V IV explained in terms of the known chemistry of Mo , Mo , and Mo . Quanti tative investigation would be difficult, however, because of the complexity of the phenomena involved. The empirical rate law obtained, for example, was based simply on determining total molybdenum remaining in solution after filtration of slurry samples. It is of fractional order in both reactants and reflects the behaviour at one pH only. The overall process of reduction and precipitation probably involves a combination of consecutive and parallel reactions as well as the complex equilibria attendant with the - 78 -involvement of polymeric molybdate and mixed valent species. The activation energy obtained of 14.1 kcal/mole represents that of the overall reduction and precipitation process. It does indicate at least that chemical pro cesses are rate controlling. ) The gravimetric, oxidimetric, and thermogravimetric results are in i fair agreement with those expected if the Mo precipitate wa's MoCKOH)^. It was found that on storage in a dessicator the precipitate oxidized slightly and this may explain the low oxidimetric results. The difference of the gravimetric and thermogravimetric results from the expected values is probably due to variations of the storage and drying procedure before weighing samples for analysis. Since no chloride ion was found in the precipitates analyzed it is concluded that the sodium present was the result of the ion exchange process described by Souchay et al.5"'" The increase of sodium content for precipitation carried out in 3 M NaCl is consistent with this interpretation. The observed stoichiometry can be interpreted in terms of the pro duction of nitrogen and ammonia according to the reaction N2H4 »• JjN + NH3 + e~ (1) and the production of hydroxylamine according to the reaction 2H20 + N2H"4 >• 2NH2OH + 2H+ + 2e~ (2) -3 It was found that consumption of 6.95 x 10 moles of hydrazine generated 1.616 x 10 ^ moles of nitrogen. Assuming all the nitrogen was generated according to equation 1 and that the balance of the hydrazine reacted according to equation 2 yields the following results: - 79 -NH3 produced = 3.232 x 10 moles -3 NH^OH produced = 7.436 x 10 moles -3 electrons generated = 10.668 x 10 moles The stoichiometry with respect to reduction of MoVI to MoV obtained is thus 1.53 which is in good agreement with the average stoichiometry found of 1.55. Since the volume of solution used was 200 ml the resultant concentrations of ammonia and hydroxylamine would be 0.016 M and 0.037 M respectively. Both of these concentrations are small enough to escape detection by the analytical techniques used. The stoichiometry observed in the present work is lower than that 79 778283 reported in the literature which varies between 2 and 4. ' ' None of the results reported in the literature, however, were obtained under condi tions of pH, concentration, and mole ratio of reactants similar to. those of the present work. Given the variable nature of the proposed hydrazine oxidation reactions and the possibility of different molybdenum species acting as oxidants it is difficult to relate results obtained under one set of conditions to those obtained under another. There does not seem to be any reason not to suppose that hydrazine could react to produce hydroxylamine in a 2 electron path and ammonia and nitrogen in a 1 electron path. It thus seems likely that one of the unidentified reaction products mentioned by Ostrowetsky and Brinon was ammonia. From the point of view of the present study, however, the important fact is that the economically favourable 4 electron stoichiometry is not obtained in the conditions of interest. - 80 -7. CONCLUSION It is apparent that neither hydrogen reduction nor reduction with hydrazine is an ideal method for recovery of molybdenum from hypochlorite leach solutions. Both reductants require acidification of the solution to be treated so there is no advantage to be gained over solvent extraction in this respect. Hydrogen reduction can only be carried out at a reasonable rate in rather severe conditions from a materials standpoint and an efficient catalyst is required. Reduction with hydrazine involves a low stoichiometry plus the require ment of an appreciable excess of hydrazine over the stoichiometric amount to obtain a reasonable rate of precipitation. The precipitate produced by hydrazine reduction will contain on the order of 3% sodium and copper must be eliminated from the leach solution or it will contaminate the precipitate. In addition the rate of precipitation is significantly decreased in the 3 M Nadexpected in the leach solution. The barren solution from precipitation by hydrazine would contain at least 0.05 g/1 molybdenum plus excess hydrazine. Recycle of this solution for^hypochlorite regeneration would require investi gation of the effect of hydrazine and molybdate on the electrolytic process employed. Reduction with hydrazine could, however, be carried out at a relative ly low temperature and in a simple reactor. One way to perform the reaction would be to add hydrazine and acid to the leach solution in a small stirred tank and then discharge the reacting mixture to a thickener sized to give a suitable residence time. Precipitate could be discharged as a slurry and the overflow recycled for hypochlorite regeneration. The discharge slurry - 81 -could be filtered for precipitate recovery. It is clear that the cost of hydrazine in relation to the price of molybdenum and the marketability of an MoO(OH) product contaminated with sodium are the deciding factors in the feasibility of using hydrazine for molybdenum recovery. On the basis of this study recovery of 1 kg of molyb denum contained in MoO(OH)3 would require consumption of about 1.2 kg of hydrazine to obtain reasonable kinetics and completeness of precipitation. At the current price for hydrazine of $3.52/kg in tank car lots it would cost $4.22 for hydrazine per kg of molybdenum as MoO(OH) . Molybdenum oxide is currently selling for approximately $20/kg contained molybdenum. At the current price of $20/kg Mo as molybdic acid and $3.52/kg hydrazine in tank car lots it would cost $4.22 for hydrazine per kg Mo contained in MoO(OH) . - 82 -8. REFERENCES 1. Sutolov, Alexander. Copper porphyries. Salt Lake City, Utah, University of Utah Printing Services, 1974. 2. Warren, I.H. et al. Canadian Institute of Mining and Metallurgy. CIM annual volume, 1977, p. 11. 3. Lindstrom, R.D. and B.J. Scheiner. U.S. Bureau of Mines report of investigations 7802, 1974. 4. Fischer, D.D. et al. U.S. Bureau of Mines report of investigations 8088, 1975. 5. Barr, D.S. et al. International journal of mineral processing, v. 2, p. 303, 1975. 6. Ibid. v. 4, p. 83, 1977. 7. Scheiner, B.J. et al. U.S. Bureau of Mines report of investigations 8145, 1976. 8. Scheiner, B.J. et al. Extraction and recovery of molybdenum and rhenium from molybdenite concentrates by electrooxidation: prototype cell demon stration. Paper presented at the 16th annual conference of metallur gists, sponsored by the Metallurgical Society of CIM, Vancouver, B.C., 1977. 9. Mounsey, Diana M. 10. Lyapina, Z.M. and A.N. Zelikman. Tsvetnye metally, 1959, p. 93. 11. Zelikman, A.N. and Z.M. Lyapina. Tsvetnye metally, 1960, p. 119. 12. Sobol, S.I. Gosudarstvennyi nauchno-issledovatel1skii institut tsvetnykh metallov, Moscow. Sbornik nauchnykh trudov, v. 18, p. 414, 1961. 13. Zelikman, A,N, and Z.M. Lyapina. Planseeberichte fur pulvermetallurgie, v. 8, p. 148, 1961. 14. Wagenmann, Rolf. East German Patent 57,595, 1967. 15. Kunda, W. and B. Rudyk. Planseeberichte fur pulvermetallurgie, v. 13, p. 157, 1965. 16. Messner, Martin E. and Stuart R. Zimmerly. U.S. Patent 3,376,104, 1968. - 83 -17. Lindsay, D.G. Endako roasting practice. Paper presented at the 16th annual conference of metallurgists sponsored by the Metallurgical Society of CIM, Vancouver, B.C., 1977. 18. Pourbaix, Marcel. Atlas of electrochemical equilibria in aqueous solutions, 2nd English ed. Houston, National Association of Corrosion Engineers, 1974. 19. Sasaki, Yukiyoshi and L.G. Sillen. Arkiv for kemi, v. 29, p. 253, 1967. 20. Cotton, Frank Albert and Beoffrey Wilkinson. Advanced inorganic chemistry: a comprehensive text, 3rd ed. New York, Interscience, 1972. 21. Rollinson, Carl L. in Comprehensive inorganic chemistry. Oxford, Pergamon, 1973, v. 3, p. 623. 22. Baes, Charles F. and Robert E. Messner. The hydrolysis of cations. New York, Wiley, 1976. 23. Schwing, Jean-Paul. Journal de chimie physique, v. 61, p. 508, 1964. 24. Aveston, J. et al. Inorganic chemistry, v. 3, p. 375, 1964. 25. Sasaki, Yukiyoshi and L.G. Sillen. Acta chemica scandinavica, v. 18, p. 1014, 1964. 26. Sillen, L.G. Pure and applied chemistry, v. 17, p. 55, 1968. 27. Schwarzenbach, G. and J. Meier. Journal of inorganic and nuclear chemistry, v. 8, p. 302, 1958. 28. Honig, Dan S. and Kenneth Kustin. Inorganic chemistry, v. 11, p. 65, 1972. 29. Collin, Jean-Paul et al. International conference on the chemistry and uses of molybdenum, 1st Reading, England, 1973. Proceedings. Edited by P.C.H. Mitchell. London, Climax Molybdenum Co. Ltd., 1973, p. 59. 30. Chariot, Gaston. L'analyse quantitative et les reactions en solution, 4th edition. Paris, Masson, 1957. 31. Holtje, R. and R. Geyer. Zeitschrift fur anorganische und allgemeine chemie, v. 246, p. 243, 1941. 32. Grasshoff, Klaus and Harry Hahn. Zeitschrift fur analytische chemie, v. 186, p. 132, 1962. - 84 -33. Schwing, Jean-Paul. Academie des Sciences, Paris. Comptes rendu, v. 254, p. 4018, 1962. 34. Zelikman, A.N. Molibden. Moscow, Metallurgiya, 1970. 35. Ardon, Michael and Arnold Pernick. Journal of the less common metals, v. 54, p. 233, 1977. 36. Viossat, B. and M. Lamache. Societe chimique de franee, Paris. Buttetin part 1, 1975, p. 1570. 37. Clausen, Donald F. and John H. Shroyer. Analytical chemistry, v. 20, p. 925, 1948. 38. Glemser, Oskar and Gertrud Lutz. Zeitschrift fur anorganische und allgemeine chemie, v. 264, p. 17, 1951. 39. Sacconi, Luigi and Renato Cini. Journal of chemical physics, v. 18, p. 1124, 1950. 40. Treadwell, W.D. and Y. Schappi. Helvetia chimica acta, v. 29, p. 771, 1946. 41. Arnold, R. and Sheila M. Walker. South african chemical institute. Journal, v. 9, p. 80, 1956. 42. Ostrowetcky, Simone. Societe chimique de franee, Paris. Bulletin, 1964, p. 1003. 43. Malprade, L. in Nouveau traite de chimie mineral. Edited by Paul Pascal. Paris, Masson, v. XIII. 44. Weiser, Harry B. The hydrous oxides. New York, McGraw-Hill, 1926. 45. Filippov, M.P. and Ya A. Nuger. Russian journal of inorganic chemistry, v. 10, p. 148, 1965. 46. Palmer, . Experimental inorganic chemistry. Cambridge, Cambridge University Press, 1954. 47. Mellor, Joseph William. A comprehensive treatise on inorganic and theoretical chemistry. New York, Longman, Green and Co. 1922 - [37], v. 11, p. 525. 48. Simon, J.P. and P. Souchay. Societe chimique de franee, Paris. Bulletin 1956, p. 1402. 49. Ardon, M. and A. Pernick. Inorganic chemistry, v. 12, p. 2484, 1973. - 85 -50. Katsobashvili, Ya R. Russian journal of inorganic chemistry, v. 5, p. 1295, 1960. 51. Souchay, Pierre et al. Societe chimique de franee, Paris. Bulletin 1970, p. 892. 52. Haight, G.P. Acta chemica scandinavica, v. 15, p. 2012, 1961. 53. Bergh, A.A. and G.P. Haight. Inorganic chemistry, v. 1, p. 688, 1962. 54. Guibe, L. and P. Souchay. Journal de chemie physique, 1957, p. 684. 55. Souchay, Pierre et al. Academie des sciences, Paris. Comptes rendu serie C, v. 262, p. 1524, 1966. 56. Ardon, Michael and Arnold Pernick. American chemical society. Journal, v. 95, p. 6871, 1973. 57. Ardon, Michael et al. American chemical society. Journal, v. 98, p. 2338, 1976. 58. Chalilpoyil, Purush and Fred C. Anson. Inorganic chemistry, v. 17, p. 2418, 1978. 59. Lagrange, Phillippe and J.P. Schwing. Academie des sciences, Paris. Comptes rendu serie C, v. 263, p. 848, 1966. 60. Lagrange Phillippe and Jean-Paul Schwing. Societe chimique de franee, Paris. Bulletin 1968, p. 536. 61. Lamache-Duhameaux, M. et al. Journal de chemie physique, v. 65, p. 1921, 1968. 62. Smith, Edgar F. American chemical journal, v. 1, p. 329, 1879-80. 63. Smith, Edgar F. and W.S. Hoskinson. American chemical journal, v. 7, p. 90, 1885. 64. Watt, George W. and Darwin D. Davies. American chemical society. Journal, v. 70, p. 3751, 1948. 65. Samartsev, A.G. and E.I. Levitina. Zhurnal fizicheskoi khimii, v. 32, p. 1023, 1958. 66. Wherry, Edgar T. and Edgar F. Smith. American journal of chemistry, v. 29, p. 806, 1907. 67. Paal, C. and G. Brunjes. Deutsche chemische gesellschaft. Berichte v. 47, p. 2214, 1914. - 86 -68. Paal, C. and Hans Buttner. Deutsche chemische gesellschaft. Berichte, v. 48, p. 220, 1915. 69. Barner, Herbert E. and Richard V. Scheuerman. Handbook of thermo-chemical data for compounds and aqueous species. New York, Wiley, 1978. 70. Browne, A.W. and F.F. Shetterly. American chemical society. Journal, v. 31, p. 783, 1909. 71. Bray, William C. and Eustace J. Cuy. American chemical society. Journal, v. 46, p. 1796, 1924. 72. Kirk, R.E. and A.W. Browne. American chemical society. Journal, v. 50, p. 337, 1928. 73. Higginson, W.C.E. et al. Chemical soceity, London. Journal, 1953, p. 1388. 74. Higginson, W.C.E. and D. Sutton. Chemical society, London. Journal, 1953, p. 1402. 75. Cahn, John W. and Richard E. Powell. American chemical society. Journal, v. 76, p. 2568, 1954. 76. Higginson, W.C.E. in Chemical society, London. Special publication 10 "Recent aspects of the inorganic chemistry of nitrogen," 1957, p. 95. 77. Jokob, W.F. and W. Kozlowski. Roczniki chemii, v. 9, p. 667, 1929. 78. Rao, G. Gopala and M. Suryanarayana. Zeitschrift fur analytische chemie, v. 168, p. 177, 1959. 79. Ostrowetsky, Simone and Danielle Brinon. Academie des sciences, Paris. Comptes rendu serie C, v. 263, p. 406, 1966. 80. Chauveau, Francoise et al. Academie des sciences, Paris. Comptes rendu, v. 240, p. 194, 1955. 81. Audrieth,, L.F. and B.A. Ogg. The chemistry of hydrazine. New York, Wiley, 1951. 82. Huang, T. and J.T. Spence. Journal of physical chemistry, v. 72, p. 4198, 1968. 83. Mishra, H.C. and R.N.P. Sinha. Indian journal of chemistry, v. 9, p. 1300, 1971. 84. Furman, N. Howell and W.M. Murray. American chemical society. Journal, v. 58, p. 1689, 1936. - 87 -85. Vogel, Arthur I. A text-book of quantitative inorganic analysis. London, Longman, Green and Co., 1949, p. 447. 86. Ismay, Arnaldo Andres. Selective leaching of molybdenum from mixed copper-molybdenum sulfides. M.A.Sc. thesis. University of British Columbia, 1976. 87. DeVries, John E. and E. St. Clair Gantz. Analytical chemistry v. 25, p. 973, 1953. 88. Wardlaw, William and Norman Darby Sylvester. Chemical society, London. Journal, p. 969, 1923. 89. Young, Roland S. Chemical analysis in extractive metallurgy. London, Griffin, 1971. 90. Callahan, Clarence M. et al. Analytical chemistry, v. 32, p. 635, 1960. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items