The influence of calcium, magnesium, phosphorus and silica on vanadium precipitation in acidic sulfate solutions

The influence of calcium, magnesium, phosphorus and silica on vanadium precipitation in acidic sulfate solutions

A. H. Martins

Universidade Federal de Minas Gerais, Departamento de Engenharia Metalúrgica e de Materiais

Rua Espírito Santo 35, sala 206 - Belo Horizonte - MG, Brazil, CEP 30160-030

Phone: [phone], Fax: [phone]

E-mail: [e-mail]

(Received: September 15, 1999 ; Accepted: April 15, 2000)

Abstract - The experimental research work was carried out in accordance with the Factorial Design Statistical Method to evaluate and analyze the influence of calcium, magnesium, silica and phosphorus on the precipitation of vanadates. Precipitation was performed by neutralization with H₂SO₄ of alkaline aqueous solutions containing vanadium pentoxide (V₂O₅) at 60± 2ºC. The experimental responses measured were percent of vanadium recovery and vanadium pentoxide content in the precipitate. These impurities are considered representative of those present in a leach liquor from the ore obtained at Campo Alegre de Lourdes (Brazil). The operational variables in this work were used under optimum conditions as determined by the statistical approach.

Among the impurities under study, phosphorus exhibited the highest negative influence on the experimental responses. Phosphorus diminished the percent of vanadium recovered from 98.9 to 34.5 and the vanadium oxide content in the precipitate from 91.2 % to 39.3 %.

Keywords: vanadium, precipitation, hydrometallurgy

INTRODUCTION

Precipitation of inorganic vanadium compounds from leach liquors containing impurities is a very difficult hydrometallurgical operation to perform. This is because the chemical behavior of vanadium in aqueous solution follows the unpredictable pattern of the transition elements in many of its compounds. This causes uncertainty about the precise reaction in the acid and alkali media and metal concentrations commonly used in industrial plants. Vanadium solubilized in aqueous solution may present different oxidation states ( II, III, IV and V) as anionic and cationic complexes, as initially reported by Ducret (1945), Rossotti and Rossotti (1956) and Pope and Dale (1968). This characteristic makes precipitation of inorganic vanadium compounds a difficult hydrometallurgical task, especially when metallic ionic impurities are present in the aqueous solution. These impurities may change the precipitation characteristics and the chemical properties of the vanadate precipitated.

These precise reactions are very important when the goal is to precipitate vanadium compounds used in catalyst production. Vanadium pentoxide and several different inorganic vanadates are widely used for preparation of catalysts that have an important role in the chemical industry. One example of a catalyst based on vanadium compounds is the one used in sulfuric acid production by the contact process. Taking into account industrial growth and its demands for materials with well-defined chemical, physical and morphological characteristics, the study of the precipitation of vanadium inorganic compounds in aqueous solutions having typical leach liquor impurities became a demanding technological task (Martins, 1988; Bento, 1991; Bento, 1995; Martins, 1997).

The main goal of this work was to study the influence of some impurities typical of leach liquors on the precipitation of vanadium compounds. In this work a statistical approach to the experimental design using the Factorial Method (Cox, 1958; Duckworth, 1968; Box, Hunter and Hunter, 1978) was adopted, permitting us to evaluate the influence of its impurity and interactions on the recovery of vanadium from the precipitation solution and the vanadium oxide content in the vanadate precipitate. Moreover, it allowed evaluation of the relationship between the precipitate and the individual or combined influence of impurities on the chemical composition of the precipitate. Thus, it would be possible to determine how much vanadium is recovered from the aqueous solution and the vanadium pentoxide content in the precipitate as well, for sulfate acid precipitation solutions containing calcium, magnesium, phosphorus and silica as impurities.

The main experimental research work was divided into two parts: i) study of the influence of those operational variables considered most important in the precipitation of inorganic vanadium compounds based on vanadium recovery and vanadium oxide content in the precipitate as experimental responses (Martins, 1999) and ii) study of the influence of calcium, magnesium, phosphorus and silica on vanadium precipitation. Such impurities are found in typical leach liquors of a titaniferrous-magnetite Brazilian ore bearing vanadium (Campo Alegre de Lourdes). This paper reports the results obtained for the study of impurities on the vanadium precipitation.

EXPERIMENTAL PROCEDURE

The aqueous solutions used in all tests were 0.4-M vanadium prepared by solubilizing suitable amounts of V₂O₅ in 2.0 N NaOH solution. All chemicals used were reagent grade.

The chemical composition of the standard solution of vanadium used in this work was the same as that found in most of the leach liquors of ores bearing vanadium, especially in the hydrometallurgical treatment of a titaniferrous-magnetite Brazilian ore (Campo Alegre de Lourdes) bearing vanadium (Martins, 1988; Bento, 1991; Bento, 1995; Martins, 1997).

Vanadium solutions were titrated with H₂SO₄ solutions in a cylindrical 100-mL glass beaker. The solutions were titrated with 2.0-mL per unit of time to reach equilibrium, i.e., the time necessary to stabilize the pH measurement of the aqueous solution after each addition of titration solution. Once equilibrium was reached, it was assumed that the end of the protonation reactions for vanadium species in aqueous solution. The beaker was immersed in a constant temperature bath and mixed by a 4.0-cm diameter glass propeller with three circular blades and driven by a mechanical device. The temperature of the solution was 60± 2ºC. Initial pH for each test solution was determined by a digital pH meter, properly calibrated every hour during the experiment with regular buffer solutions.

The impurities used in this work were added to 100-mL of the vanadium aqueous solution as Ca (OH) ₂, Mg (OH) _2, Na₂HPO₃ .12H₂O and Na₂O.SiO₂ .All these chemicals were also reagent grade.

The optimum experimental conditions previously determined by the statistical approach used in the tests (Martins, 1999) were: a 3.0 N H₂SO₄ concentration in the titration solution, a stirring intensity of 120 rpm, a vanadium solution prepared before each test (considered new solution), 2.0-mL per unit of time of titration solution added to the precipitation solution bearing vanadium to reach equilibrium as the rate of neutralization, an initial pH of 1.7 for precipitation and a temperature of 60± 2ºC for the system. There was no addition of seed for precipitation.

The vanadate precipitates were filtered and dried in an oven at 120± 2ºC for 30 minutes. The amounts of vanadium and impurities in solution and in the precipitate were determined by the atomic absorption spectrophotometry methodology (Polikarpova and Panteleeva, 1983) using Perkin-Elmer model 5000. Identification of inorganic compounds obtained by the precipitation procedure was carried out using the Phillips x-ray diffractometry equipment, model PW1130/60.

A statistical design for technological experiments using the Factorial Method was adopted (Cox, 1958; Duckworth, 1968; Box, Hunter and Hunter, 1978). This allowed us to reduce the number of experiments required to precipitate the vanadium without jeopardizing the quality of the technological information obtained.

Table 1 presents the four impurities under study and their numeric values adopted for two experimental levels. They were used in the Factorial Design Method to plan the tests and evaluate the influence of impurities on vanadate precipitation.

Table 2 shows the matrix block for 16 replicated experiments in accordance with the Factorial Method for four variables (A, B, C, D) at upper (+) and lower (-) levels. The experimental responses replicated for each test were vanadium recovery from the precipitation solution and percentage of vanadium oxide in the precipitate. The notation column was included to facilitate identification of the tests where there was individual (A, B, C, D) or interaction (AB, AC, BC, ABC, AD, BD, ABD, CD, ACD, BCD, ABCD) influence of the variables on the experimental response. For example, in test number 1, all variables were at their lower experimental level, while in test number 2 only variable A was at the upper level and the others were at their lower level.

The percentage of vanadium recovery is an important experimental response because it is related to vanadium removal from leach liquors for future disposal. Vanadium is one of the most poisonous elements, and its presence in aqueous solutions or solid industrial residues is a very worrisome hydrometallurgical problem for preservation of the environment. It is also related to the efficiency of the precipitation process, while the vanadium content in the precipitate assumes an important role in catalyst performance for several reactions producing strategic compounds for the chemical industry.

Vanadium oxide content is particularly important for catalyst production because the content of vanadium pentoxide is related to catalyst performance in several reactions that are very important to the chemical industry, such as sulfuric acid production, synthesis of special printing inks, cracking of oil and synthetic plastics, among others.

RESULTS AND DISCUSSION

Table 3 shows the replicated experimental responses based on the Statistical Factorial Method for vanadate precipitation, where R1 and R2 were the percentage of vanadium recovery from the precipitation solution and R3 and R4 were the percentage of vanadium oxide in the precipitate.

Table 4 shows a statistical analysis of the experimental results for the influence of impurities on vanadium recovery from the precipitation solution. These numerical values were obtained from some statistical parameters such as statistical confidence interval of 95% in the results, 16 statistical degrees of freedom and 1.74 Student's t-test multiplier (Cox, 1958; Duckworth, 1968; Box, Hunter and Hunter, 1978).

In Table 4, the individual influence of variables A (Ca), B (Mg), C (P) and D (SiO₂) and their interactions, AB, BC, ABC, AD, BD, ABD, CD, ACD, BCD and ABCD on vanadium recovery were statistically significant at a confidence level of 95 %. The AC interaction was not significant; i.e there was no statistical difference between experimental levels when the interaction AC changed from the lower level to the upper one. This means that the phosphorus concentration in the precipitation solution was high enough to inhibit the effect of calcium on vanadium recovery.

Based on the results shown in Table 4, phosphorus presented the highest influence on the vanadium recovered from the precipitation solution, and this influence was negative. This means that an increase in phosphorus concentration in solution causes a decrease in the percentage of vanadium recovery. The highest percentage of vanadium recovery ranged from 52.7 to 96.5 % for tests without phosphorus in the precipitation solution (see Table 3, R1 and R2). Thus, the presence of phosphorus caused the precipitation of vanadates with a low percentage of vanadium and this has a negative effect on catalyst performance, for instance. Thermodynamic calculations performed by Ogasawara and Silva (1990) showed that phosphorus in the precipitation solution reduces the vanadate ions activity, making it more difficult for precipitation reactions to occur. It can be concluded that purification of the aqueous solution for precipitation of vanadium is required to remove phosphorus from the precipitation solution.

Magnesium showed the second highest influence on vanadium recovery. It also presented a negative effect, i.e., when the magnesium concentration in the precipitation was increased, vanadium recovery decreased. Calcium, exhibited almost the same influence as that of the magnesium, had the third highest effect.

Decreasing the pH of the system, leads to an increase in the activity of Ca²⁺ and Mg²⁺ ions (Ogasawara and Silva, 1990). This benefits the precipitation of calcium and magnesium sulfate at a pH value lower than that of vanadate precipitation. Thus, Ca²⁺ and Mg²⁺ ions act as competitors of vanadium ions for precipitation. Nevertheless, the concentration of calcium and magnesium should be reduced to minimum values in solution to increase vanadium recovery.

Silica showed the least influence on vanadium recovery, but the influence was positive. This means that the presence of silica in the precipitation solution compensated for the negative influence of the other elements. As a matter of fact, the positive contribution of silica was very small compared to the negative effect of phosphorus.

Analysis of the interactions of the impurities on vanadium recovery, showed a positive effect when calcium or magnesium was added to a solution containing phosphorus. Test number seven in Table 4 showed a positive effect on vanadium recovery for a simultaneous increase in the magnesium and phosphorus concentrations in the precipitation solution. This means that magnesium and calcium had a kind of neutralizing effect on the behavior of phosphorus in solution.

The contribution of interaction AC to vanadium recovery was considered insignificant because its statistical confidence level was 75%, much lower than the confidence level initially established for this work (95%). This means that there was no difference between the upper and lower experimental levels for the interaction to affect the expected response.

Applying Fisher's Method (Cox, 1958; Duckworth, 1968; Box, Hunter and Hunter, 1978) to the Factorial Design permitted us to determine an algebraic expression representing the statistical model adopted in this work. This statistical model was used to evaluate the influence of the impurities on vanadium recovery (R'). The algebraic expression is R' = 51.30 - 3.50 X_A - 5.15 X_B + 3.05 X_AX_B - 15.85 X_C + 1.95 X_BX_C - 2.60 X_AX_B X_C + 2.10 X_D + 1.50 X_AX_D + 2.15 X_BX_D - 1.80 X_AX_BX_D + 3.40 X_CX_D - 1.25 X_AX_CX_D -3.10 X_BX_CX_D + 2.20 X_AX_BX_CX_D. This expression assumed an important role in this study because it permitted us to predict the percentage of vanadium recovery from the precipitation system poisoned with Ca, Mg, P and SiO₂, by attributing values -1 and +1 to the lower and upper levels of the X parameters of each variable, respectively. Table 5 shows the predicted response for vanadium recovery, using the algebraic expression of the statistical model and the experimental response obtained on a laboratory bench scale.

From a comparison between the experimental results and the statistically predicted responses, it can be concluded that the statistical approach may provide percentages of vanadium recovery in the vanadate precipitation under the experimental conditions studied.

Table 6 presents a statistical analysis of the experimental results for the influence of impurities on the percentage of vanadium oxide in the precipitate. These numerical values were obtained from the following statistical parameters: a statistical confidence interval of 95% in the results, 16 statistical degrees of freedom and 1.74 Student's t-test multiplier (Cox, 1958; Duckworth, 1968; Box, Hunter and Hunter, 1978).

Phosphorus showed the greatest influence on vanadium oxide content in the vanadate precipitate, and it also exhibited a negative effect. This means that an increase in phosphorus concentration in solution causes a decrease in the percentage of vanadium oxide in the precipitate. The highest percentage of vanadium oxide content in the precipitate ranged from 65.7% to 91.2% for tests without phosphorus in the precipitation solution (see Table 3, columns R3 and R4). The presence of phosphorus, even in small concentrations, inhibits vanadate precipitation that has a vanadium oxide content acceptable for further application in catalysis.

The second and third highest influences on the vanadium oxide in the precipitate were magnesium and calcium, respectively. They also presented a negative effect, i.e., when the magnesium and calcium concentrations in the precipitation solution were increased vanadium oxide precipitation decreased.

Silica presented the least influence on vanadium recovery, and its behavior was similar to that shown for the percentage of vanadium recovery discussed earlier.

Analyzing the influence of the interaction between impurities on the vanadium oxide content in the precipitate, interactions AC, BC, AD, BD, ABD, ACD, and ABCD were considered insignificant because their statistical confidence level ranged from 75% to 50%, and these values are much lower than 95%, the value established for this experimental work.

The algebraic expression for the statistical model also determined by application of Fisher's Method to the Factorial Design permitted prediction of the percentage of vanadium oxide content in the precipitate (R''). The expression is R'' = 57.45 - 3.55 X_A - 4.38 X_B + 1.66 X_AX_B - 17.63 X_C + 0.3 X_AX_C + 0.5 X_BX_C - 1.12 X_AX_B X_C + 2.86 X_D + 0.64 X_AX_D + 0.26 X_BX_D - 0.04 X_AX_BX_D + 3.25 X_CX_D - 0.52 X_AX_CX_D + 1.34 X_BX_CX_D + 0.41 X_AX_BX_CX_D.

As mentioned previously, this expression permitted the prediction of the percentage of vanadium pentoxide content in the precipitate by the statistical approach, attributing +1 and-1 values to the upper and lower levels of the X parameters of each variable, respectively. It can be noted that the interactions considered did not contribute significantly with very small numerical values for the response.

Table 7 shows the predicted response for the vanadium oxide content in the precipitate, using the statistical approach and the experimental response obtained. The numerical values obtained for the predicted response showed only a slight difference from those in experimental results. The high number of insignificant interactions identified by the statistical model can explain this. These interactions increased the error associated with the algebraic expression. However, the predicted responses had numerical values very close to the experimental values.

Table 8 shows the inorganic compounds precipitated from the vanadium alkaline solution containing a specific impurity, as determined by the X-ray diffraction method. The inorganic compounds presented in order of decreasing radiation intensity.

Analysis of the inorganic compounds shown in Table VIII indicates that decavanadate salt (V₁₀O₂₈ ) was present in the precipitate as the vanadium salt with the highest relative concentration in the pure system and aqueous solutions containing impurities as well. Vanadium oxide (V₂O₅.3H₂O) was the species with the second highest concentration for the system with Mg, P or SiO₂.

The systems containing Mg or SiO₂ were responsible for generating a precipitate with another decavanadate species form (V₁₀O₂₄) as the third highest concentration compound. On the other hand, calcium enhanced the precipitation of other vanadate forms less commonly used as catalyst.

It is important to remark that for some catalytic purposes, decavanadates and vanadium pentoxide compounds are considered the most desirable because the vanadium content is higher than that of the other vanadium compounds. Thus, vanadium alkaline solutions containing impurities such as magnesium and silica should be treated to keep them with in a specific range of concentrations in solution that allows to precipitate decavanadate salts and vanadium pentoxide. Therefore, it is possible to conclude that the approach used in the precipitation procedure was able to produce vanadium salts with chemical compositions suitable for catalytic applications.

CONCLUSIONS

Based on the experimental results and a statistical analysis, it was possible to reach the following main conclusions using these optimum conditions: a 3.0 N H₂SO₄ concentration in the titration solution, a stirring intensity of 120 rpm, preparation of vanadium solution before each test (considered new solution), 2.0-mL per unit of time of titration solution to the precipitation solution bearing vanadium to achieve equilibrium as the rate of neutralization, an initial pH of 1.7 for precipitation and a temperature of 60± 2ºC for the system.

(a) among the impurities under study, phosphorus exhibited the highest negative influence on the experimental responses, reducing the percentage of vanadium recovery from 98.9 to 34.5, For the percentage of V₂O₅ content in the precipitate, the reduction was from 91.2 to 39.3;

(b) All impurities studied in this work decreased the percentage of vanadium recovery and vanadium pentoxide content in the precipitate;

(c) The presence of magnesium and silica permitted us to obtain precipitates with higher concentrations of decavanadate salts and vanadium pentoxide. These compounds are considered to be the most desirable for catalytic purposes.

ACKNOWLEDGEMENTS

The authors are grateful to PRONEX-MCT-Brazil, Coordenação de Aperfeiçoamento de Pessoal de Ensino Superior CAPES and Conselho Nacional de Desenvolvimento Científico e Tecnológico CNPq Brazil for their financial support.

• Bento, C.A., A Precipitaçăo de Vanadatos de Sódio em Meio Cloreto, M.Sc. Thesis, Faculdade de Engenharia Química de Lorena- Programa de Pós-Graduaçăo em Engenharia de Materiais, Săo Paulo, Brazil, 64 p. (1991) (in portuguese).
• Bento, C.A. and Martins, Comparaçăo entre os Sistemas Sulfato e Cloreto Visando a Precipitaçăo de Vanadatos, A.H., XVI Encontro Nacional de Tratamento de Minérios e Hidrometalurgia, 3-6 / Setembro, Rio de Janeiro, Brazil, pp. 136-151 (1995) (in portuguese).
• Box, G.E.P., Hunter, W.G. and Hunter, J.S., Statistics for Experimenters, John Wiley & Sons Inc., New York (1978).
• Cox, D.R., Planning of Experiments, John Wiley & Sons Inc., New York (1958).
• Duckworth, W.E., Statistical Techniques in Technological Research, Methuen & Co. Ltd., London (1968).
• Ducret, L.P., Contribution a L'Etude des Ions des Valences Quatre et Cinc du Vanadium, Ann.Chimie, 6, pp. 705-773 (1945).
• Martins, A.H., A Precipitaçăo de Compostos de Vanádio, Ph.D. Thesis, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil (1988) (in portuguese).
• Martins, A.H., Some Aspects on Vanadates Precipitation, 5th Southern Hemisphere Meeting on Mineral Technology, Ed. Instituto de Tecnologia Minera, Buenos Aires, Argentina, May 6-9, pp. 181-184 (1997).
• Martins, A.H., Vanadium Precipitation from Sulfate Acid Solutions, Canadian Metallurgical Quarterly, 39, no. 1, pp. 15-22 (1999).
• Polikarpova, N. and Panteleeva, E. Yu., Atomic Absorption Determination of Vanadium in Products of the Metallurgical Industry and Mineral Ores, E. Yu., Plenum Publishing Co., translated from Zavodskaya Laboratoya, 49, nş4, April, pp. 37-40 (1983).
• Pope, M.T. and Dale, B.W., Isopoly-Vanadates Niobates and Tantalates, Quarterly Review, 22, pp. 527-548 (1968).
• Rossotti, F.J.C. and Rossotti, H., Isopolyvanadates in Acidic Solution, J. Inorg. Nucl. Chem., 2, pp. 201-202 (1956).
• Ogasawara, T. and Silva, F.T., Desenvolvimento de Diagramas Log C - pH do Sistema V-Na- H₂O a 298,15K, XIV Encontro Nacional de Tratamento de Minérios e Hidrometalurgia, Ed. Associaçăo Brasileira de Metais, Brasil, Setembro 9-12, pp. 935-949 (1990) (in portuguese).

Publication Dates

• Publication in this collection
  06 July 2000
• Date of issue
  June 2000

History

• Received
  15 Sept 1999
• Accepted
  15 Apr 2000