Production of Metallic Vanadium by Preform Reduction Process

Akihiko Miyauchi1;* and Toru H. Okabe2

1Department of Materials Engineering, Graduate School of Engineering, the University of Tokyo, Tokyo 113-8656, Japan2Institute of Industrial Science, the University of Tokyo, Tokyo 153-8505, Japan

A fundamental study was conducted on a new process for producing vanadium (V) metal by the preform reduction process (PRP) based onmetallothermic reduction of vanadium pentoxide (V2O5). Feed preforms with good mechanical strength even at elevated temperatures wereprepared by adding either CaO or MgO to V2O5 feed powder because V2O5 has a low melting point of 963K; thus complex oxides (CaxVyOz,MgxVyOz) with high melting point at more than 1273K were obtained. Reduction experiments were conducted by using either Ca or Mg vaporat 1273K for 6 h. V metal with a purity of more than 99% was successfully obtained when using Mg as a reductant. The feasibility of producingV metal by the PRP will be discussed on the basis of fundamental experiments. [doi:10.2320/matertrans.M2010027]

(Received January 26, 2010; Accepted March 16, 2010; Published April 28, 2010)

Keywords: vanadium, reduction process, metallothermic reduction

1. Introduction

Vanadium (V) is a transition-metal element with an atomicnumber of 23. V metal is the 20th most abundant elementamong all elements in the earth’s crust. The abundance of Vmetal is 120 ppm in the earth’s crust, which is significantlylarger than the well-used common metals such as nickel (Ni,84 ppm, 23rd rank) and copper (Cu, 60 ppm, 26th rank).1,2)

However, V metal belongs to a set of less common metals or‘‘rare metals’’ since the production volume of V metal wasonly 58 kt in 2007 and is far smaller as compared to that ofcommon metals.3) The small production volume of V metal ispartly due to its low concentration in the ore and unevendistribution of its minerals. The principal V metal mineralssuch as titanomagnetite contain only 1–2mass% of V2O5.

4)

The majority of the mineral resources are distributed in threecountries: China, Russia, and South Africa.

V metal and its alloys are mainly used as additive elements(alloying elements) or as catalysts. In Japan, more than 85%of V metal is used as an additive in steel products forimproving their tensile strength and heat resistance.5) Specialsteel containing V metal is applied in bridges, industrialtools, etc. In the field of chemical industries, V compoundsare utilized as a desulfurization catalyst in sulfuric acidproduction processes. Furthermore, vanadium-titanium (V-Ti) alloys are now attracting considerable attention as newelectrode materials in hydrogen storage batteries, because Vhas high hydrogen storage capacity at ambient temperatureand moderate pressure.6–8) Considering the expanding marketfor hydrogen storage batteries, the demand for V-Ti alloysmay be expected to increase in the future.

Currently, V feed for smelting V metal is produced in theform of its oxide (V2O5) as a byproduct of steel slag or as aresidue of oil, and these V2O5 feeds are mostly used as astarting material for V products. V metal is commerciallyproduced by the aluminothermic reduction (ATR) ofV2O5.

9,10) Although this process is simple and economical,the product is not a high-purity V metal; it is a vanadium-aluminum (V-Al) alloy containing 20mass% of aluminum

(Al). In order to produce high-purity V metal, multiplemelting steps by using an electron beam melting process arenecessary for removing the Al. For this reason, the ATRprocess is not a suitable production process of high-purity Vmetal. In the past, some researchers attempted to developan alternative to the ATR process for producing V metal.Metallic V with more than 99% purity was first producedthrough calciothermic reduction by Marden and Rich.11)

They used a mixture of V2O5, Ca and CaCl2 at 1173–1223K.McKechenie and Seybolt proposed another reaction based oncalciothermic reduction with a small amount of flux suchas CaI2.

12) Gregory researched calciothermic reduction ofV2O3.

13) In order to eliminate oxygen contamination,chloride metallurgy for producing high purity V metal wasinvestigated by Campbell et al.14) Preparation of high-purityV by the Van Arkel-de Boer process (Iodide disproportio-naton process) was studied by Carlson et al.15,16) However, aneffective production process suitable for commercial massproduction has not been established at this stage. Therefore,the development of a simple and efficient production processof high-purity V metal is strongly required. In the recentyears, Suzuki et al. investigated electrochemical reduction ofV2O5 or V2O3 in molten CaCl2 for producing high purity Vmetal and its alloys,17–19) as well as calciothermic reductionof V2O5 and TiO2 for producing V-Ti alloys.20) With thesebackgrounds, the present study aims to develop a new processfor the effective production of V metal from its oxide byutilizing a simple metallothermic reduction.

2. Preform Reduction Process (PRP)

As an effective production process of high-purity V metalfrom V2O5, we selected the preform reduction process(PRP).21–23) Feed preforms are prepared by mixing a startingmaterial oxide feed, flux, and binder solution. After drying,the preforms have sufficient mechanical strength to standalone. By utilizing the features of the mechanical structure ofthe feed preforms, contamination from the reduction contain-er can be prevented. In the PRP, the oxide in the preforms isdirectly reduced to metal by reductant vapor. After reduction,the produced metal is recovered by leaching. The advantages*Graduate Student, the University of Tokyo

Materials Transactions, Vol. 51, No. 6 (2010) pp. 1102 to 1108#2010 The Japan Institute of Metals

of this process are its effective control of purity andmorphology and its ability to flexibly scale the reductionprocess. Contamination into the product can be easilyavoided because the feed material in the self-supportingpreform does not have physical contact with the reactioncontainer or the reductant. Highly flexible scalability isachieved as it is possible to simultaneously treat multiplepieces of preforms in a single reduction chamber. Further-more, the amount of molten salt (or flux) used in this processis smaller than that used in other processes.

3. Thermodynamic Analysis

A reductant suitable for the PRP and the optimumreduction temperature were studied from the thermodynamicview-point. Calcium (Ca) or magnesium (Mg) was selectedas a suitable reductant for the V2O5 reduction by consideringthe Gibbs energies for the formation of oxides shown inFig. 1.24) For reducing V2O5 by the PRP, the reductant metalhas to be supplied in a vapor form. The vapor pressure ofsome selected elements as a function of temperature isillustrated in Fig. 2.24) In order to have an efficient supply of areducing agent in the gas phase, it is desirable that the vaporpressure of the reducing agent is higher than 10�4 atm at thereduction temperature.25) In this study, 1273K was selectedas the reduction temperature because their vapor pressuresare higher than 10�3 atm, which is high enough for sufficientgas supply of the reducing agents.

The melting point of V2O5 is 963K, which is lower thanthe reduction temperature. This low melting point causesdifficulty in the fabrication of a suitable feed preform for thePRP. The preforms fabricated only from V2O5 and a bindercannot retain their mechanical shapes at elevated temper-

atures at above melting point of V2O5. In order to solve thisproblem, V2O5 was mixed with flux such as magnesiumoxide (MgO) or calcium oxide (CaO) and calcined tosynthesize complex vanadium oxides. In Fig. 3, equilibriumphase diagrams of V2O5-MgO and V2O5-CaO quasi binarysystems are shown.26,27) It is expected that Mg2V2O7 andCa2V2O7 maintain enough mechanical strength even at thereduction temperature (Tred. ¼ 1273K) because these com-pounds have high melting points.

4. Experimental

A flowchart of the production of V metal by the PRPemployed in this study is shown in Fig. 4. The PRP of V2O5

using Ca or Mg vapor as a reductant consisted of four majorsteps: preform fabrication, calcination, reduction by reduc-tant vapor, and sample recovery by acid leaching. Firstly, thefeed preform was fabricated from a slurry made of V2O5

powder (99.9% purity), CaO or MgO powder as a flux, and acollodion solution (5mass% nitrocellulose in ethanol andether) as a binder. The experimental conditions for preparingthe feed preform using the Ca or Mg vapor were determined.The viscosity of the slurry was controlled by varying theamount of the flux and binder. The cationic molar ratio,RCat./V, listed in Table 1 is defined as RCat./V ¼ NCat.=NV,where NCat. and NV are the mole amounts of the cation inthe flux and V, respectively. Plate-shaped preforms with athickness of 3–10mm were prepared by casting the ob-tained slurry into a stainless steel mold. Secondly, the castpreforms were heated and calcined in air for 2 h at calcinedtemperatures.

Temperature, T / K

-1400

-1200

-1000

-800

-600

-400

-200

0

300 500 700 900 1100 1300Sta

ndar

d G

ibbs

ene

rgy

of fo

rmat

ion,

∆G

° f/ k

J m

ol-1

2 Mg + O2= 2 MgO2 Ca + O2

= 2 CaO

3/2 Fe + O2= 1/2 Fe3O4

4/3 Al + O2= 2/3 Al2O3

Ti + O2= TiO2

4/5 V + O2= 2/5 V2O5

Reduction temperature

Fig. 1 Ellingham diagram of some selected oxides.24)

2000 1000 500

Temperature, T / K

-10

-8

-6

-4

-2

Vap

or p

ress

ure

of e

lem

ent i

, log

pi

/ atm

Reciprocal temperature, 1000 T -1 / K-1

0.5 1.0 1.5 2.0

Ca

MgAl

V

Ti

Fe

P °Mg = 0.458 atmat 1273 K

P °Ca = 0.018 atmat 1273 K

Fig. 2 Vapor pressure of some selected elements as a function of reciprocal

temperature.24)

Production of Metallic Vanadium by Preform Reduction Process 1103

Figure 5(a) is a schematic illustration of the experimentalapparatus used for the calcination step. In this calcinationstep, the binder and water in the feed preforms were removed,and an adequate amount of mechanical strength was furthergranted to the preforms. Then, the calcined preforms wereinstalled in a reaction vessel and reduced by the Ca or Mgvapor in the reduction step. Figure 5(b) is a schematicillustration of the experimental apparatus used for producingV metal by the PRP. Four pieces of the calcined preforms(4–6 g each) containing vanadium oxides were placed in athick-walled stainless steel vessel, and the reductant solidwas placed at the bottom of the vessel. Two times excessstoichiometric amount of reductant (Ca or Mg) was used inthe reduction experiment (see RR values in Table 2). Thereductant solid was physically isolated from the feed pre-form, and the reductant vapor was supplied to the feedpreform. Sponge titanium was also placed at the bottom ofthe vessel for gettering the nitrogen gas in the system.After being sealed by tungsten inert gas (TIG) welding, the

Mg2V2O7 + L

Mg3(VO4)2 + L

V2O5 + L

MgV2O6 +

Mg2V2O7

V2O5 + Mg2V6O17

Mg2V6O17 + L

Mg 2

V6O

17

MgV

2O6

Mg 2

V2O

7

Mg 3

V2O

8M

g 3V

2O8+

MgO

Mg 2

V2O

7+

Mg 3

V2O

8

670°C

604°C

742°C

980°C

1074°C

MgO content, xMgO (mol%)

Tem

pera

ture

, T’

/ °C

V2O5 (MgO)20 40 600

500

700

900

1100

MgV2O6 + L

640°C

Reduction temp. in this study (1273 K)

L

Tem

pera

ture

, T’

/ °C

L

618°C

1015°C

1380°C

778°C

CaV

2O6

Ca 2

V2O

7

Ca 3

V2O

8

V2O5 (CaO)CaO content, xCaO (mol%)

600

800

1000

1200

1400

200

Reduction temp. in this study (1273 K)

(b)

(a)

40 60

80

Fig. 3 Equilibrium phase diagram of (a) V2O5-MgO quasi-binary sys-

tem,26) (b) V2O5-CaO quasi-binary system.27)

Table 1 Experimental conditions for the preparation of feed preforms.

Exp.Mass of sample, wi/g Cationic Calcined Calcined

Note

no. Feed Flux Binder molar ratio, temperature, time,(Corresponding figures)

V2O5 MgO CaO Collodion�1 RCat./V�2 Tcal./K t00cal./h

A-1 2.68 1.19 — 4.02 1.0 1173 2 Fig. 6(a)

A-2 2.68 1.79 — 4.56 1.5 1173 2

A-3 2.68 — 1.67 4.66 1.0 1173 2

A-4 2.68 — 2.48 5.25 1.5 1173 2

A-5 2.68 1.19 — 4.09 1.0 873 ! 1173 2

A-6 2.68 1.79 — 4.60 1.5 873 ! 1173 2 Fig. 6(b), Fig. 7(a) (c), Fig. 8

A-7 2.68 — 1.67 4.76 1.0 873 ! 1173 2

A-8 2.68 — 2.48 5.29 1.5 873 ! 1173 2 Fig. 7(b) (d), Fig. 9

�1Collodion solution (5mass% nitrocellulose in ethanol and ether) was used.�2RCat./V ¼ xCat.=xV, xcat.: mole amounts of cations in flux, xV: mole amounts of vanadium.

Mixing / casting

Reduced preform

V metal powder

Preform fabrication

Feed preform

Calcination

Reduction

Leaching

Vacuum drying

S L

50% CH3COOH aq.,20% HCl aq.,

Isopropanol,Distilled water,

Acetone

Waste solution

Reductant (Ca or Mg)

V2O5 BinderFlux

Fig. 4 Flowchart of the production of V metal by PRP in this study.

1104 A. Miyauchi and T. H. Okabe

steel reaction vessel was then heated in an electric furnacemaintained at a constant temperature of 1273K for 6 h; thepreforms reacted with the reductant vapor.

After 6 h of reaction, the reaction vessel was removed fromthe furnace and was quenched in water. Finally, the preformsin the container were mechanically recovered at roomtemperature by opening the sealed vessel with a lathe. Theproduct in the preforms obtained after the reduction experi-ment was recovered by leaching the preforms with an acid,i.e. by dissolving the reaction product (CaO/MgO), flux,and excess reductant in an acetic acid solution. The obtainedproduct was rinsed with hydrochloric acid at room temper-

ature. It was then successively rinsed with distilled water,alcohol, and acetone, followed by drying in vacuum. Thephases in the sample were identified using X-ray diffractionanalysis (XRD; Rigaku, Rint 2000, Cu–K� line). Thecomposition of the sample was determined by X-rayfluorescence spectrometry (XRF; JEOL Ltd., JSX-3210).

5. Results and Discussion

5.1 Calcination processThe plate-shaped preform was fabricated by casting feed

slurry prepared by mixing V2O5, flux, and the collodionsolution. The size of the preform was 50� 20� 4mm, andit was beige in color. Figure 6 shows representative photo-graphs of the obtained samples after calcination experiment.As shown in Fig. 6(a), samples (Exp. A-1-Exp. A-4) did notsustain their original shapes, after the calcination process at aconstant temperature of 1173K for 2 h because of meltingV2O5 in the feed preforms immediately after startingcalcination procedure. Samples with a cationic molar ratioof RCat./V ¼ 1:5 (Exp. A-6 and Exp. A-8) maintained theiroriginal shape (see Fig. 6(b)), after the calcination step attemperatures ranging from 873K to 1173K in 2 h. Incontrast, other samples with RCat./V ¼ 1:0 (Exp. A-5 andExp. A-7) did not retain their original shapes since they losttheir mechanical strengths due to the melting of V2O5.The preforms which retained their shapes were successfullyproduced in the calcination process by adjusting both theamount of flux and the calcination temperature. Figure 7presents the appearance and X-ray diffraction patterns ofthe samples (Exp. A-6 and Exp. A-8) obtained after thecalcination process. The shape and color of the preform wereretained after performing calcination at temperatures risingfrom 873K to 1173K in 2 h. By the reaction between V2O5

and the flux, Mg2V2O7 or Ca2V2O7 were formed in theobtained preforms. Although the melting point of Mg2V2O7

is slightly lower than 1273K (see Fig. 3), the preformsdemonstrated sufficient mechanical strength even at elevatedtemperatures. This is probably because MgO and Mg3V2O8

in the preforms contribute to maintaining the shape ofpreforms, and Mg2V2O7 is reduced to V metal or suboxidesbefore melting the parts of preforms. The detail reason isunder investigation.

(b)

(a)

Stainless steel cover

Stainless steel plate

Reductant (Ca or Mg)

Stainless steel reaction vessel

Ti sponge getter

Feed preform (V2O5 + flux)

TIG weld

Ceramics chamber

Thermocouple

Heating element

Alumina crucible

Feed preform

Alumina tube

Fig. 5 Schematic illustration of the experimental setup for (a) the

calcination experiment, (b) the reduction experiment.

Table 2 Experimental conditions for the reduction experiments, and representative analytical results of vanadium powder obtained after

reduction.

Cationic ExcessMass of samples, wi/g

Composition of vanadium powder

Exp. molar reductantReduction Reduction

obtained after reduction, Ci�3/mass%

no.Flux Reductant

ratio, ratio,temperature, time, Calcined

Reductant,RCat./V

�1 RR�2 Tred./K t00red./h preform,

wR/gCV CMg CCa CFe CCr

wcal./g

A-6-1 MgO Mg 1.5 2.0 1273 6 3.937 3.345 99.7 0.2 — 0.01 0.03

A-6-2 MgO Ca 1.5 2.0 1273 6 3.670 4.906 86.0 2.4 10.6 0.3 0.5

A-8-1 CaO Mg 1.5 2.0 1273 6 3.474 2.573 85.4 — 13.0 0.2 0.4

A-8-2 CaO Ca 1.5 2.0 1273 6 3.604 3.976 79.0 — 20.4 0.1 0.4

�1RCat./V ¼ xCat.=xV, xcat.: mole amounts of cations in flux, xV: mole amounts of vanadium�2Excess reductant ratio: RR ¼ wR=wR,theo., wR,theo.: stoichiometic mass of reductant necessary for reduction.�3Determined by XRF; value excludes carbon and gaseous elements.

Production of Metallic Vanadium by Preform Reduction Process 1105

5.2 Magnesiothermic reduction (Exp. A-6-1 and Exp. A-8-1, reductant: Mg)

Figure 8 illustrates the appearance and the representativeXRD patterns of the samples with MgO flux obtained aftereach step in Exp. A-6-1, and the experimental conditions forreduction experiments are presented in Table 2. The colorof the preform changed to a jet-black color after the PRPby Mg vapor at 1273K for 6 h (Fig. 8(a)), and the shape ofthe preform was slightly deformed. However, it was easy torecover the preform from the reaction vessel because it wasphysically isolated from the vessel even after reduction.These results indicate that this PRP is suitable for avoidingcontamination from the reaction vessel. During the leachingstep in the acetic acid solution, the original shape of thepreform was lost and a powder with a grayish black colorwas obtained without pulverizing the reduced preform(Fig. 8(b)). Figure 8(c) presents an XRD pattern of thepreform after reduction at 1273K for 6 h. All the complexoxides were reduced to V metal by the Mg vapor. Themorphology of the sample was sponge like metal powder. Asummary of the analytical results of the obtained V powder

after reduction as well as of other samples is listed in Table 2.The XRF analysis revealed the purity of the obtainedvanadium powder to be 99.7mass% V, 0.2mass% Mg, and0.03mass% Cr (Exp. A-6-1). In contrast, the preform withCaO flux (Exp. A-8-1) was not completely reduced to Vmetal. The XRD analysis revealed that the black powderobtained after leaching process consisted of V metal andCaV2O4. This is probably because the calcium complex

(b) Solid preform

(a) Melted preform

Fig. 6 Photograph of the obtained samples in the alumina crucible after the

calcination experiment, (a) Exp. A-1 (Flux: MgO, RCat./V ¼ 1:0, Tcal. ¼1173K), (b) Exp. A-6 (Flux: MgO, RCat./V ¼ 1:5, Tcal. ¼ 873 ! 1173K).

(c) Exp. A-6 (Flux: MgO, aRCat. / V = 1.5)

(a) Exp. A-6 (Flux: MgO, aRCat. / V = 1.5)

(b) Exp. A-8 (Flux: CaO, aRCat. / V = 1.5)

(d) Exp. A-8 (Flux: CaO, aRCat. / V = 1.5)

Mg2V2O7 JCPDS # 70-1163

Inte

nsity

, I (

a.u.

) MgO JCPDS # 77-2364

20 30 40 50 60 70 80Angle, 2 θ (degree)

Ca2V2O7 JCPDS # 72-2312

CaO JCPDS # 77-2010

20 30 40 50 60 70 80Angle, 2 θ (degree)

Inte

nsity

, I (

a.u.

)

Fig. 7 The obtained sample after the calcination process. (aRCat./V ¼xCat.=xV, xcat.: mole amounts of cations in flux, xV: mole amounts of

vanadium.) (a) Photograph of the preform (Exp. A-6, Flux: MgO,

RCat./V ¼ 1:5, Tcal. ¼ 873 ! 1173K), (b) Photograph of the preform

(Exp. A-8, Flux: CaO, RCat./V ¼ 1:5, Tcal. ¼ 873 ! 1173K), (c) X-ray

diffraction pattern of the sample (Exp. A-6), (d) X-ray diffraction pattern

of the sample (Exp. A-8).

1106 A. Miyauchi and T. H. Okabe

oxides (i.e. CaV2O4) formed in the preform are chemicallystable and it becomes difficult to reduce by Mg vapor. Thedetails are unclear at this stage.

5.3 Calciothermic reduction (Exp. A-6-2 and Exp. A-8-2, reductant: Ca)

Figure 9 presents the appearance and the representativeXRD patterns of the samples obtained after each step inExp. A-8-2 with CaO flux. When Ca was employed as thereductant, black solid in the form of flake was obtained after

leaching. All of Ca2V2O7 was reduced to CaV2O4 with theCa vapor after reduction, that is, the oxidation state of Vchanged from +v to +iii. However, metallic V was notproduced in the experimental conditions employed in thisstudy. In the same way, V metal was not produced in Exp. A-6-2, in which MgO was used as a flux; rough black powderobtained after leaching was identified as Mg1:5VO4 (i.e.Mg3V2O8) by XRD analysis. The XRF analysis revealed thepurity of the obtained powder in the Exp. A-8-2 and Exp. A-6-2 to be higher than 85mass% V. It is assumed that a part of

(c) After reduction

(a) After reduction

(b) After leaching

(d) After leaching

V : JCPDS # 22-1058

Inte

nsity

, I (

a.u.

)

MgO : JCPDS # 77-2364

Angle, 2θ (degree)30 40 50 60 70 80

V : JCPDS # 22-1058

Angle, 2θ (degree)30 40 50 60 70 80

Inte

nsity

, I (

a.u.

)

Fig. 8 Photograph and XRD pattern of the obtained sample (Exp. A-6-1),

(a) and (c): After the reduction process (R: Mg, Flux: MgO, Tred. ¼1273K, t00red. ¼ 6 h), (b) and (d): After the leaching process (50%

CH3COOH aq. (t00 lea. ¼ 12 h), 20% HCl aq. (t00 lea. ¼ 1 h)).

(c) After reduction

(a) After reduction

(b) After leaching

(d) After leaching

Inte

nsity

, I (

a.u.

)

CaV2O4 JCPDS # 74-1359

CaO JCPDS # 77-2010

30 40 50 60 70 80

CaV2O4 JCPDS # 74-1359

Angle, 2θ (degree)

30 40 50 60 70 80Angle, 2θ (degree)

Inte

nsity

, I (

a.u.

)

Fig. 9 Photograph and XRD pattern of the obtained sample (Exp. A-8-2),

(a) and (c): After the reduction process, (R: Ca, Flux: CaO, Tred. ¼ 1273K,

t00red. ¼ 6 h), (b) and (d): After the leaching process (50% CH3COOH aq.

(t00 lea. ¼ 12 h), 20% HCl aq. (t00 lea. ¼ 1 h)).

Production of Metallic Vanadium by Preform Reduction Process 1107

the surface of the obtained powder was reduced to metallic V.At this stage, the reason for the unsuccessful results ofcalciothermic reduction is not clear. The difference betweenthe results obtained using the Mg and Ca reductants areprobably due to the difference in their vapor pressures. Thevapor pressure of Mg (0.458 atm) is 26 times larger than thatof Ca (0.018 atm) at 1273K. In fact, the amount of Careductant remained in the crucible after the reduction waslarger than that of Mg. From a thermodynamic view-point(see Fig. 1), Ca is a more favorable and strong reductantcompared to Mg. Therefore, certain kinetic reasons may havehindered the calciothermic reduction of vanadium oxidesunder PRP employed in this study, e.g., the slow diffusion ofCa in the CaO byproduct formed at the surface of thereduction product, and/or formation of compounds acting asa kinetic barrier.

6. Conclusion

In order to develop a new reduction process for producingfine vanadium powder using the metallothermic reduction,the preform reduction process (PRP) has been applied.Slurries obtained by mixing V2O5 powder, a flux (e.g. MgO,CaO) and a binder (e.g. collodion), were cast into molds andthen dried to obtain the preforms. Before the reductionprocess, fabricated preforms with RCat./V ¼ 1:5were calcinedby heating from 873 to 1173K in 2 h in order to producethe preforms with good mechanical strength at elevatedtemperatures. The sintered solid preforms containing tocomplex oxides (MgxVyOz, CaxVyOz) were then reactedwith Ca or Mg vapor at a constant temperature of 1273K for6 h. When Mg vapor was used as a reductant, pure vanadiumpowder of more than 99mass% purity was obtained afterleaching process. The PRP was thus demonstrated to besuitable for producing a fine, homogeneous metallic vana-dium powder.

Acknowledgements

The authors are grateful to Profs. M. Maeda, S.Yamaguchi, Y. Mitsuda, and K. Morita of The Universityof Tokyo for their generous support and valuable discussions,and Messrs. K. Tomoda and Y. Watanabe of the NipponCatalyst Cycle Co. Ltd. for providing valuable technicalinformation. We would specially like to thank Dr. K. Yasudaand Mr. K. Yanada for their useful discussions. Thanks are

also due to Mr. T. Oi for providing technical assistance inproducing the metal powder.

REFERENCES

1) The chemical society of Japan: Chemical Handbook Basic Edition of

the Fourth Edition, (Maruzen Co. Ltd., Tokyo, 2002).

2) P. Enghag: Encyclopedia of the Elements, Trans. by T. Watanabe,

et al., (Asakura Publishing Co. Ltd., Tokyo, 2007) pp. 295–304.

3) USGS, 2006. Mineral Industry Surveys, US Dept. of the Interior, US

Geological survey, Mar. 2008.

4) F. Habashi: Handbook of extractive metallurgy III, (WILLEY-VCH,

Weinheim, 1997) pp. 1471–1488.

5) Japan Oil, Gas and Metals National Corporation, Mineral Resources

Information Center: Metallic Resources Report, Mar. (2007).

6) H. Tamura: Hydrogen Storage Alloys—Fundamentals and Frontier

Technologies—, (NTS Inc., Tokyo, 1998).

7) M. Tsukahara, K. Takahashi, T. Mishima, A. Isomura and T. Sakai:

J. Alloy. Compd. 224 (1995) 133–138.

8) E. Ronnebro, D. Noreus, T. Sakai and M. Tsukahara: J. Alloy. Compd.

231 (1995) 90–94.

9) H. Yoshinaga: J. Min. and Mater. Process. Inst. Japan 123 (2007) 768–

771.

10) T. K. Mukherjee and C. K. Gupta: J. Less-Common Metals 27 (1972)

251–254.

11) J. W. Marden and M. N. Rich: Vanadium Ind. Eng. Chem. 19 (1927)

786–788.

12) R. K. McKechenie and A. U. Seybolt: J. Electrochem. Soc. 97 (1950)

311–315.

13) E. D. Gregory: J. Electrochem. Soc. 98 (1951) 395–399.

14) T. T. Campbell, J. L. Schaller and F. E. Block: Metal. Trans. 4 (1973)

237–241.

15) O. N. Carlson and C. V. Owen: J. Electrochem. Soc. 108 (1961) 88–92.

16) J. W. Nash, H. R. Ogden, R. E. Durtschi and I. E. Campbell: J.

Electrochem. Soc. 100 (1953) 272–275.

17) R. O. Suzuki and H. Ishikawa: ESC Trans. 3 (2007) 347–356.

18) R. O. Suzuki and H. Ishikawa: Proc. Sohn Int. Symp., Advanced

Proceeding of Metals and Materials, 3 ed. by F. Kongoli and R. G.

Reddy, (The Minerals, Metals &Materials Society, 2006) pp. 597–607.

19) Y. Oka and R. O. Suzuki: J. Japan Inst. Metals 72 (2008) 181–187.

20) R. O. Suzuki, K. Tatemoto and H. Kitagawa: J. Alloy. Compd. 385

(2004) 173–180.

21) T. H. Okabe, T. Oda and Y. Mitsuda: J. Alloy. Compd. 364 (2004) 156–

163.

22) T. H. Okabe, S. Iwata, M. Imagunbai and M. Maeda: ISIJ Int. 43 (2003)

1882–1889.

23) B. Yuan and T. H. Okabe: J. Alloy. Compd. 443 (2007) 71–80.

24) I. Barin: Thermochemical Data of Pure Substances, 3rd ed., (VCH

Verlagsgesellschaft mbH, Weinheim, Germany, 1995).

25) T. Azakami and Y. Awakura: Extractive Metallurgy, (The Japan

Institute of Metals, Tokyo, 1999).

26) R. C. Kerby and J. R. Wilson: Can. J. Chem. 51 (1973) 1032–1040.

27) A. N. Morozov: System CaO-V2O5, Metallurg (Leningrad) 13 (1938)

21–28.

1108 A. Miyauchi and T. H. Okabe