Quantitative Reduction of Iron under Nitrogen Atmosphere for Potassium Dichromate Titration

Abstract

Total iron contents in iron ores have been accurately determined by JIS M 8212, in which iron ions in digested solutions of iron ores are reduced to divalent prior to redox titration. It is necessary for the iron reduction process that no reducing chemicals other than iron(II) in the decomposition solutions must not remain after the reduction with titanium(III). However, the redox reactions concerning the chemical species present in the decomposition solution has not been completely elucidated at the present time. In this paper, the redox reactions that occurred in the decomposition solution during the iron reduction in JIS M 8212 were studied by potentiometry and spectrophotometry under nitrogen atmosphere. The redox reaction of tin(II)/(IV) was very slow, causing significant effects on identifying the end point of the indicator for the iron reduction. The copper chloro-complexes were reduced with titanium(III) at a potential higher than that of indigo carmine used as a redox indicator, so that the reduced copper(I) gave a positive error to the potassium dichromate titration. The pentavalent vanadium was reduced with titanium (III) to form a complex with titanium, which also interfered with the potassium dichromate titration positively. To avoid these interferences, titanium(III) chloride was stoichiometrically added to the reaction mixture after addition of tin(II) chloride under nitrogen atmosphere so as to reduce only iron to divalent prior to the following redox titration. Combination of the proposed protocol with the potassium dichromate titration could successfully determine the iron content of certified reference materials of iron ores.

1. Introduction

Accurate determination of the iron content in iron ores has been of great importance for their quality assurance. The standard methods for determining the iron content in iron ores used in Japan and internationally are JIS M 8212¹⁾ and ISO 9507,²⁾ respectively. The typical protocols adopted by both standard methods are as follows: After the iron ores are digested with a mixture of mineral acids, the valence of the dissolved iron ions is strictly set to two by using a reductant, such as titanium trichloride. The digested solution is then subjected to a redox titration with potassium permanganate. To obtain accurate analytical results, this method requires strict reduction, in which divalent iron ions are the only chemical species that possess reductive activity in the digested solution.

Titanium(III) chloride,³⁾ tin(II) chloride,⁴⁾ hydrogen sulfide, and silver⁵⁾ have been used as reductants in the reduction process in JIS M 8212. The reduction method using a combination of titanium(III) chloride and tin(II) chloride was studied in detail by Saeki et al.³⁾ and is the basis of the current JIS M 8212. Prior to Saeki’s study, excess tin(II) used to be added to reduce all the iron species to divalent forms in the digested solution.⁴⁾ This method was replaced by other methods because of the utilization of mercury(II) chloride to deactivate the excess tin(II). Mercury compounds are usually avoided in every standard chemical analysis method. In methods using silver powder,⁵⁾ iron is reduced to a divalent state via a redox reaction between metallic silver and iron(III). This method was adopted from ISO 9508. Aluminum^6,7,8,9) has also been intensively investigated as a reductant, although this method was not adopted in JIS M 8212. Because Al(III) ions do not have other valence states, the Al(III) ions resulting from the redox reaction do not disturb the redox reactions of iron(II) and chromium(VI). The redox reaction between aluminum and iron(II) requires 2.3–3.2 mol/L of hydrochloric acid. However, because the dilution of the digested solution is unavoidable for the subsequent redox titration, this reduction method has not been included in the standard method. Potassium borohydride is also a potential reductant because iron(III) in a solution prepared by digesting iron ores with sulfuric acid is effectively reduced by potassium borohydride.¹⁰⁾ However, the reduction with potassium borohydride is severely hindered by copper when hydrochloric acid is present in the digestion solution. Thus, because hydrochloric acid is used to prepare the digested solution of iron ores, reduction with potassium borohydride has not been adopted in JIS M 8212.

In the reduction of iron with titanium(III) chloride and tin(II) chloride adopted in JIS M 8212, most of the iron(III) is first reduced by tin(II) chloride, and then the remaining iron species are completely reduced by the addition of excess titanium(III) chloride. After the completion of the iron reduction, the remaining excess titanium(III) is deactivated into a trivalent state using potassium dichromate. This preliminary reduction step is performed under atmospheric conditions. Dissolved oxygen in the atmosphere interferes with iron reduction. In addition, other redox-active species in the digested solutions of iron ores interact not only with the iron species but also with each other, which results in a complicated redox system related to the iron species. Thus, experience and training are required to accurately determine the endpoints based on the discoloration of indigo carmine as a redox indicator. If we can quantitatively monitor the redox reactions involving the metal species in the digested solution during the preliminary reduction step, accurate and quantitative reduction of the iron species to the divalent state can be achieved without the need for experience or skill.

Measurement of the redox potential and absorbance of solutions containing redox-active and colored species are direct and effective approaches for quantitatively monitoring redox reactions in solutions. However, to the best of our knowledge, redox reactions involving metal species in digestion solutions of iron ores have not been investigated using potentiometry and spectrophotometry. In this study, we monitored the redox reactions occurring in digested solutions of iron ore by measuring the redox potential and absorbance of the solutions. To obtain high reproducibility and accuracy, the experiments were conducted under a nitrogen atmosphere. Precise experiments revealed the following: Tin(II), which was initially added to reduce the iron species in the digested solution of iron ores, was oxidized to tin(IV) via a redox reaction with iron(III). This tin(IV) was reduced to tin(II) by adding excess titanium(III) in the subsequent step. Because these redox reactions of tin are extremely slow at room temperature, unintended errors in determining the end points of the reduction may occur in the subsequent potassium dichromate titration. In addition, copper and vanadium interfered with the accurate determination of the endpoint in the subsequent potassium dichromate titration. When copper coexists in a digested solution containing hydrochloric acid, the chloro-complex of copper provides a redox buffering zone above the redox potential of indigo carmine, causing a positive error in the determination of the endpoint. The chloro-complex of copper(II) consumed titanium(III) to become a monovalent chloro-complex during the reduction by titanium(III). The formed copper(I) chloro-complex caused a positive error in the subsequent potassium dichromate titration. Titanium(III) added to the digestion solution reduced vanadium(V) to vanadium(III), which was used in the subsequent potassium dichromate titration and caused a positive error. To overcome these errors, we devised a novel protocol in which titanium(III) chloride, which was used as a reducing agent, was added until the reduction of iron was completed by measuring the redox potential of the solution. The resulting solution was titrated immediately with potassium dichromate. The analytical results of the iron concentration in an aqueous hydrochloric acid solution containing iron(III) obtained using both devised protocols agreed with each other. Furthermore, when the protocol was applied to certified standard iron ore materials, the analytical results were identical to the certified values.

2. Experiment

2.1. Reagents and Apparatus

Hydrochloric acid, sulfuric acid, phosphoric acid, iron(III) chloride hexahydrate, titanium(III) chloride solution, potassium dichromate, indigo carmine, sodium diphenylamine-4-sulfonate, sodium hydroxide, vanadyl sulfate hydrate, and copper(II) nitrate trihydrate were purchased from Kanto Chemical Co. (Tokyo, Japan). Analytical-grade tin (II) chloride dihydrate, disodium ethylenediaminetetraacetate tetrahydrate, sodium acetate trihydrate, granular metallic tin, and ammonium metavanadate were purchased from Wako Pure Chemical Industries (Tokyo, Japan). Tap water was purified using a combination of Millipore Elix Advantage 5 and Elga PURELAB Ultra Ultrapure Water Production Systems. Only purified water (18 MΩ or higher) was used in the experiment.

Certified reference materials of iron ores, JSS 831-2, JSS 850-3, and JSS 841-1, were used and decomposed according to the protocol described in JIS M 8212 to prepare the digested solutions. Aqueous iron(III) chloride solutions that mimicked the digested solutions of iron ore reference materials were prepared for primary studies using the following procedure: 17.7 g of iron(III) chloride hexahydrate was dissolved in hydrochloric acid to prepare 1000 mL of a 6.5×10⁻² mol/L iron(III) chloride solution such that the final concentration of hydrochloric acid was 0.3 mol/L. A 0.44 mol/L tin(II) chloride solution was prepared by dissolving 10.0 g of tin(II) chloride dihydrate in 20 mL of heated hydrochloric acid, and the final volume of the solution was then set to 100 mL by adding water. This solution was stored in a brown bottle in a cool, dark place after a few granules of metallic tin were added to the bottle. A 0.13 mol/L titanium(III) chloride solution was prepared by adding 6 mol/L hydrochloric acid to 1 mL of a 200 g/L titanium(III) chloride solution to make 10 mL. A 0.01667 mol/L potassium dichromate standard solution was prepared by dissolving 4.903 g of potassium dichromate in water and diluting the solution to 1000 mL by adding water.

A JASCO (Tokyo, Japan) V-650 UV-visible spectrophotometer equipped with a quartz cell and an optical path length of 1 cm and a JASCO EHC-716 Peltier cell holder was used to measure the UV-visible absorption spectra. A HORIBA (Kyoto, Japan) pH/ION METER F-73-equipped HORIBA 9300-10D ORP electrode was used to measure the pH and redox potential of the solution.

2.2. Fabrication of Iron(III) Reduction and Titration Equipment

Figures 1(a) and 1(b) show the schematics of the equipment assembled to reduce the iron(III) contained in the digested solution of the iron ore, followed by titration of the resulting solution. The equipment consisted of a magnetic stirrer with a heating plate, 200-mL beaker used as a vessel, silicone lid, Liebig condenser, ORP electrode, and nitrogen line. Nitrogen gas entered the vessel from the nitrogen line and was discharged through a condenser. The water that evaporated during the experiments was returned to the beaker through a condenser. The lid was sealed with tape to maintain airtightness. Furthermore, a hole on the lid was used not only for chemical addition but also for bullet insertion. The hole was closed with a glass stopper when not required. The open end of the condenser was narrowed to prevent air from entering.

Fig. 1. Apparatus of redox titration under nitrogen atmosphere.

2.3. Reduction of Iron(III) and Redox Titration with Potassium Dichromate

Figure 2 shows the scheme of the proposed method for quantifying the total iron content in digested solutions of iron ore. Briefly, 70 mL of an iron(III) chloride solution or a digested solution of iron ore was placed in a vessel on the heating plate shown in Fig. 1. While nitrogen was bubbling, the solution was heated to 90°C, and 5 mL of 0.44 mol/L tin(II) chloride solution was added to it. After cooling the solution to room temperature under a nitrogen atmosphere, a 0.13 mol/L titanium(III) chloride solution was added to the solution while monitoring the potential until the reduction to iron(II) was completed (the redox potential of the solution was approximately 290 mV).

Fig. 2. Schematic procedure of proposed reduction followed by titration for analysis of total iron in digested solutions of iron oress.

The amount of Fe in the digested solution was determined using the following titration procedure: After adding 10 mL of a 0.01667 mol/L potassium dichromate standard solution to the sample solution containing iron prepared using the above procedure, 30 mL of mixed acid (sulfuric acid: phosphoric acid: water = 3:3:14, v/v/v) was added at 2 g/L. Sodium diphenylamine sulfonate solution (0.5 mL) was added, and titration was performed using a 0.01667 mol/L potassium dichromate standard solution. Changes in the redox potential were recorded during the titration.

3. Results and Discussion

3.1. Effect of Dissolved Oxygen on Stabilization of Redox Potential

Because JIS M 8212 does not specify the working atmosphere, an analytical protocol is usually performed under atmospheric conditions, which causes the dissolution of oxygen from the air into a titrating solution. Therefore, the effect of dissolved oxygen on reduction was investigated by comparing the changes in the redox potential and absorbance of the solution obtained under nitrogen and aerial conditions.

First, to study the differences in the heating conditions and atmospheres during reduction by tin(II) chloride, four experimental conditions were set, in which the heating conditions were a temperature of either 25°C or 90°C and the atmosphere was either nitrogen or air. Figure 3(A) shows the change in absorbance at 315 nm of a model solution, to which a 0.44 mol/L tin(II) chloride solution was added. The absorption spectrum at approximately 315 nm corresponds to the absorption of the chloro-complex of iron(III). Under all four experimental conditions examined, the absorbance at 315 nm decreased linearly as the amount of tin(II) chloride increased. The slopes are almost identical under all conditions. This indicates that the reduction reaction of iron(III) by tin(II) proceeded in the same manner under all experimental conditions. In other words, the reduction of iron(III) by tin(II) is not affected by dissolved oxygen, suggesting that it does not necessarily need to be carried out at a high temperature of 90°C.

Fig. 3. (A) Plots of absorbance at 315 nm of Fe(III) solution vs. amounts of SnCl₂ added under the following conditions; ■, 90°C, Aerial; ●, 90°C, Nitrogen; □, 25°C, Aerial; 〇, 25°C, Nitrogen. (B) Plots of redox potential (■) and absorbance at 315 nm (〇) of Fe(III) solution vs. amounts of SnCl₂ added at 25°C under nitrogen atmosphere. Experimental conditions: 70 mL of 0.3 mol/L hydrochloric acid solution containing 4.6×10⁻³ mol Fe(III), was used to measure redox potential. Aliquots of a 0.44 mol/L SnCl₂ solution containing 2.3 mol/L HCl were added.

Figure 3(B) shows the change in the redox potential of the mimic solution upon the addition of tin chloride under nitrogen at room temperature. The potential range in which the redox potential decreased slightly at approximately 500 mV corresponds to the potential buffering zone of Fe(II)/Fe(III). When the volume of the tin(II) chloride aqueous solution reached 5 mL, the reduction of iron(III) was complete, and the potential began to drop abruptly, as shown in Fig. 3(B). Meanwhile, the absorbance at 315 nm, assigned to the chloro-complex of iron(III), did not change.

In the reduction process described in JIS M 8212, the iron species are completely reduced by the addition of tin(II) chloride, followed by titanium(III) chloride. Therefore, the effect of dissolved oxygen on the reduction of Ti (III) chloride was investigated. A 0.13 mol/L titanium(III) chloride solution was added to 70 mL of a 0.3 mol/L hydrochloric acid solution containing 4.6×10⁻³ mol iron(III) chloride and 2.2×10⁻³ mol tin(II) chloride solution under a nitrogen atmosphere and air. This solution corresponds to the digested solution to which tin(II) chloride solution has been added, according to the procedure described in JISM8212. Figure 4(A) shows the relationship between the amounts of titanium(III) chloride added and the redox potentials measured under nitrogen and air atmospheres. When the redox potential reached approximately 300 mV under atmospheric conditions, the potential did not drop, even when a solution of titanium(III) chloride was added beyond the stoichiometric volume of the solution. This behavior is caused by the dissolved oxygen and can be explained by the redox reaction shown in Eq. (1).

  

$${\mathrm{O}}_2+4{\mathrm{H}}^{+}+4{\mathrm{e}}^{-}=2{\mathrm{H}}_2\mathrm{O}\hspace{1em}\hspace{1em}E°/\text{mV}=122.9$$(1) ¹¹⁾

Fig. 4. (A) Plots of redox potential of Fe(III) solution after addition of SnCl₂ vs. amounts of TiCl₃ added at 25°C. (B, C) Change in redox potential with time course of Fe(III) solution vs. amounts TiCl₃ added at room temperature under (B) nitrogen atmosphere and (C) aerial atmosphere. Arrows denote injection points of TiCl₃ solutions. Experimental conditions: 70 mL of 0.3 mol/L hydrochloric acid solution containing 4.6×10⁻³ mol Fe(III) and 2.2×10⁻³ mol SnCl₂ was used to measure redox potential at room temperature. Aliquots of a 0.13 mol/L TiCl₃ solution containing 6 mol/L HCl were injected.

To investigate the influence of dissolved oxygen, we monitored the changes in the redox potential over time at approximately 300 mV after the addition of titanium(III) chloride solution. In a nitrogen atmosphere, the redox potential decreased monotonically immediately after the addition of the titanium(III) chloride solution and then quickly reached a constant value, as shown in Fig. 4(B). However, in the case of the atmospheric environment, although the potential dropped rapidly immediately after the addition of the titanium(III) chloride solution, it slowly recovered, as shown in Fig. 4(C). Note that the recovery profile is not linear; the potential rises slowly for a certain period immediately after the addition of titanium(III) chloride and then rapidly in approximately 20 min. This redox behavior can be explained by the reaction between the added titanium(III) chloride and dissolved oxygen. When titanium(III) chloride is consumed by dissolved oxygen, the redox potential increases slowly. After the complete consumption of titanium (III), the potential rises rapidly to reach the redox potential buffering zone of dissolved oxygen, as described in Eq. (1). These results indicate that dissolved oxygen has a significant effect on the redox potential of the solution, and that nitrogen purging can suppress this effect.

3.2. Redox Reactions of Tin with Titanium(III) Chloride and Potassium Dichromate

As discussed in the previous section, nitrogen purging can effectively suppress the influence of dissolved oxygen on the redox reactions, allowing us to study the redox reactions that occur below 300 mV. We investigated the redox reactions of tin(II) chloride, which is initially used as a reducing agent in JIS M 8212. The redox potential of 70 mL of a 0.3 mol/L hydrochloric acid solution containing 4.6×10⁻³ mol iron(III) chloride and 2.2×10⁻³ mol tin(II) chloride solution was monitored at room temperature or 90°C under nitrogen atmosphere during the addition of a 0.13 mol/L titanium(III) chloride solution. Figure 5(A) shows the change in redox potential as a function of the amount of titanium(III) chloride solution added. When titanium(III) chloride was added in an amount greater than that required to entirely reduce iron at room temperature, the potential dropped to the buffering potential of titanium (III)/(IV). On the other hand, the addition of excess titanium(III) chloride at about 90°C lowered the solution potential to about 0 mV, which corresponds to the tin(II)/(IV) buffering potential. The redox potential did not decrease to the titanium(III)/(IV) buffering potential.

Fig. 5. (A) Plots of redox potential of Fe(III) solution after addition of SnCl₂ vs. amounts of TiCl₃ added under nitrogen atmosphere. (B, C) Change in redox potential with time course of Fe(III) solution after addition of aliqots of a TiCl₃ solution under nitrogen atmosphere at (B) 25°C and (C) 90°C. Arrows denote injection points of TiCl₃ solutions. Experimental conditions: 70 mL of 0.3 mol/L hydrochloric acid solution containing 4.6×10⁻³ mol Fe(III), 2.2×10⁻³ mol Sn was used to measure redox potential under nitrogen atmosphere. Aliquots of a 0.13 mol/L TiCl₃ solution containing 6 mol/L HCl were injected.

To elucidate the redox reactions involving tin in detail, we monitored the change in the redox potential over time after the addition of a titanium(III) chloride solution at a potential of approximately 0 mV. Figures 5(B) and 5(C) show the potential profiles obtained at 25°C and at 90°C, respectively. The redox potential at 90°C dropped rapidly immediately after the addition of titanium(III) chloride and then recovered to approximately 0 mV. This potential profile is due to the progress of the redox reaction between Ti (III) and Sn (IV). This reaction is very slow, requiring 30 min for the potential profile to reach a plateau. Saeki et al.³⁾ have also reported on the slow redox reaction of tin(IV). At 90°C, the reaction of titanium(III) with tin(IV) accelerated the redox reaction, and the redox potential profile shows a buffering zone of tin(II)/(IV) provided by the remaining tin(IV). No further drop in potential occurred because of the buffering zone. On the other hand, the reduction reaction of tin(IV) at room temperature did not proceed at 25°C due to the inertness of tin, and no buffering zones of tin(II)/(IV) were observed. Therefore, the addition of titanium(III) chloride rapidly decreased the redox potential in the buffering zone of titanium(III)/(IV).

Next, we investigated the change in the redox potential of an iron solution containing tin during oxidation with potassium dichromate. An aqueous solution of potassium dichromate was added to the iron solutions containing tin and excess titanium(III) at 90°C or at room temperature. The observed changes in the redox potential are shown in Fig. 6(A). Because the redox reaction between tin(II) and chromium(VI) progresses rapidly at 90°C, the potential rises abruptly when the oxidation of tin(II) is complete. Figure 6(B) shows the time dependence of the redox potential after the addition of potassium dichromate solution around the endpoint of tin(II) oxidation. Because the oxidation of tin(II) as well as the reduction of tin(IV) are very slow, the redox potential took approximately 30 min to reach a constant value after the addition of a potassium dichromate solution. The conditions used in this study were similar to those described in JIS M 8212. Under these conditions and in a nitrogen atmosphere, the redox potential at which the color of indigo carmine added as an indicator changed was approximately 120 mV. Therefore, the standing time required for the redox potential to exceed 120 mV and then fall below 120 mV after the addition of potassium dichromate was estimated from Fig. 6(B) to be 115 to 215 s. According to JIS M 8212, oxidation is complete when the blue color of indigo carmine is maintained for 5 s. If this criterion is followed for the oxidation treatment of the remaining titanium(III), the endpoint of the addition of potassium dichromate solution may not be possible to determine accurately. The redox reaction involving tin does not proceed readily at 25°C, on the other hand, the redox reaction between titanium(III) and chromium(VI) proceeds substantially only in the solution. As this redox reaction proceeds promptly, it took approximately 20 min for the potential to rise and fall again after the addition of potassium dichromate, as shown in Fig. 6(C). Thus, it is difficult to quickly treat the excess titanium(III) with chromium(VI) at 25°C.

Fig. 6. (A) Plots of redox potential of sample solution vs. amounts of K₂Cr₂O₇ added under nitrogen atmosphere. (B, C) Change in redox potential with time course of Fe(III) solution after addition of aliqots of a K₂Cr₂O₇ solution under nitrogen atmosphere at (B) 25°C and (C) 90°C. Arrows denote injection points of TiCl₃ solutions. Experimental conditions: 70 mL of 0.3 mol/L hydrochloric acid solution containing 4.6×10⁻³ mol Fe(III), 2.2×10⁻³ mol Sn was used to measure redox potential at room temperature. Aliquots of a 3.40×10⁻³ mol/L K₂Cr₂O₇ solution were injected.

The overall redox reactions of tin are as follows: tin(II) chloride, added as a reducing agent in JIS M 8212, was reduced from tetravalent to divalent by the subsequent addition of titanium(III) chloride, and then oxidized from divalent to tetravalent by the subsequent addition of potassium dichromate. Note that these redox reactions of tin are slow even at 90°C and are not complete within the 5 s, during which the indicator remains blue. This slow redox reaction of tin causes an erroneous determination of the endpoint of the reduction. To avoid such errors, the addition of titanium (III) chloride should be stopped when the reduction of iron species in the digestion solution of the iron ores is complete.

3.3. Reduction of Diverse Metals by Titanium Chloride

Because aluminum, bismuth, cerium, copper, manganese, nickel, lead, selenium, vanadium, and zinc are sometimes present in iron ores, the reduction of the chemical species of each element was individually studied using 70 mL of a 0.3 mol/L hydrochloric acid solution containing 4.6 × 10⁻³ mol iron(III) chloride and 2.2 × 10⁻³ mol tin(II). Figure 7 shows the typical potential changes as a function of the amount of titanium(III) chloride added. The profiles of copper(II), vanadium(V), and selenium(VI) were different from the profile of iron without any diverse elements, which was used as a control. The profiles of the other elements examined were similar to the profile of the iron control. The amount of titanium(III) chloride added for the potential to suddenly decrease differed for each element. This difference in the amount of titanium(III) chloride is due to the different amounts of tin(II) chloride added in advance; thus, the difference provides no chemical explanation.

Fig. 7. Plots of redox potential of Fe(III) solution containing a 2 mol% diverse element vs. amounts of TiCl₃ added at 25°C under nitrogen atmosphere. Experimental conditions: 70 mL of 0.3 mol/L hydrochloric acid solution containing 4.6×10⁻³ mol Fe(III), 2.2×10⁻³ mol SnCl₂ containing 9.2×10⁻⁵ mol of each diverse element except selenium was used to measure redox potential at 25°C under nitrogen atmosphere. K₂SeO₄, ZnCl₂, Cu(NO₃)₂, NH₄VO₃, and MnCl₂ were examined as a salts of diverse elements. 4.6×10⁻⁵ mol of K₂SeO₄ was added as a selenium salt. Aliquots of a 0.13 mol/L TiCl₃ solution containing 6 mol/L HCl were added.

The reduction of copper (II), selenium (VI), and vanadium (V) in the iron solution by titanium (III) chloride began immediately after the reduction of iron was complete, as shown in Fig. 7. Because the redox potential at which the indicator indigo carmine changes color is 120 mV, which is lower than the potential buffering zones of copper, selenium, and vanadium shown in Fig. 7, these three elements will be reduced to their reduced states by titanium (III) chloride if the completion of the reduction is determined only by the color change of the indigo carmine. The reduced forms of these elements cause positive errors in the subsequent potassium dichromium titration. Therefore, the reduction reactions of these elements were investigated in detail.

3.3.1. Reduction Reaction of Copper-chloro Complex

Cu(II) forms a chloro-complex with hydrochloric acid under acidic conditions. The redox reactions of hydrated copper ions and chloro-complex and their redox potentials are described in Eqs. (2)¹¹⁾ and (3).¹²⁾ The redox potentials of chloro-complexes are a function of the activity of the chloride ions. Under the conditions of 0.3 mol/L hydrochloric acid used in this study, the redox potential of the dichloro-complex is estimated to be 574.7 mV, which is 234 mV higher than that of hydrated copper ions and is close to that of iron, which is 771 mV. This is why a potential buffering zone of Cu(I)/Cu(II) was observed immediately after the reduction of iron was complete.

  

$$\mathrm{C}\mathrm{u}\left(\mathrm{I}\mathrm{I}\right)+2{\mathrm{e}}^{-}=\mathrm{C}\mathrm{u}\hspace{1em}\hspace{1em}E°/\mathrm{m}\mathrm{V}=340$$(2) ¹¹⁾

  

$$\begin{array}{l}\mathrm{C}\mathrm{u}\left(\mathrm{I}\mathrm{I}\right)+2\mathrm{C}{\mathrm{l}}^{-}+\mathrm{e}-=\mathrm{C}\mathrm{u}\mathrm{C}{{\mathrm{l}}_2}^{-}\\ E°/\mathrm{m}\mathrm{V}=513-118\mathrm{p}\mathrm{C}{\mathrm{l}}^{-}\end{array}$$(3) ¹²⁾

To confirm this, copper was added to an acidic iron solution that mimicked a digested solution of iron ore, and the redox potential of the resulting solution was monitored until it reached 100 mV or less with the addition of a titanium chloride solution. In the absence of Cu, the potential dropped quickly, regardless of the presence of Sn, as shown in Fig. 8(A). In the presence of copper, on the other hand, the potential buffering zone of Cu(I)/Cu(II) was observed, whose width was equivalent to the addition of approximately 4.3 × 10⁻⁵ mol of titanium chloride, which corresponds approximately to the amount of copper added, which was 4.6 × 10⁻⁵ mol. This indicates that the added titanium(III) chloride was used up by the reduction of copper(II). Beyond the buffering zone, the redox potential was maintained at a constant value of approximately 100 mV and did not drop further, even when excess amounts of titanium(III) chloride were added.

Fig. 8. (A) Plots of redox potential of Fe(III) (〇) and Fe(III)+ Cu(II) (□) solution vs. amounts of TiCl₃ added at 25°C under nitrogen atmosphere. (B, C) Change in redox potential with time course of a Fe(III) solution with (C) and without 9.2×10⁻⁵ mol of Cu(II) after addition of aliqots of a TiCl₃ solution under nitrogen atmosphere. Arrows denote injection points of TiCl₃ solutions. Experimental conditions: 70 mL of 0.3 mol/L hydrochloric acid solution containing 4.6×10⁻³ mol Fe(III), 2.2×10⁻³ mol SnCl₂ was used to measure redox potential at 25°C under nitrogen atmosphere. Aliquots of 0.13 mol/L TiCl₃ solutions containing 6 mol/L HCl were added.

To investigate the reasons for such a profile, we monitored the changes in the redox potential over time immediately after the addition of the titanium(III) chloride solution. Figures 8(B) and 8(C) show the changes in the potential profiles at approximately 100 mV during the addition of titanium(III) chloride solution. In the absence of copper, the potential dropped immediately upon addition of the titanium(III) chloride aqueous solution and then quickly reached a constant value, as shown in Fig. 8(B). In contrast, in the presence of copper, the potential dropped immediately after each addition of titanium chloride and then gradually increased, followed by an abrupt increase after a certain period of time, as shown in Fig. 8(C). Note that as the number of additions of a titanium chloride solution increases, the periods during which the redox potentials increase gradually after the addition of a titanium chloride solution become longer. A similar behavior of the potential was observed when titanium chloride was added to the iron solution under atmospheric conditions (Fig. 4(C)). Considering these profiles, oxygen was also involved in the recovery of the redox potential in this system. This oxygen is likely present in trace amounts as an impurity in the nitrogen gas. To verify the involvement of oxygen, titanium(III) chloride was added under atmospheric conditions to an aqueous hydrochloric acid solution containing only copper. When an aqueous titanium(III) chloride solution was added, the solution turned yellow due to the formation of a hydrogen peroxide complex of titanium(IV), which has a maximum absorption wavelength of 415 nm. The absorbance of this solution at 415 nm increased with the addition of titanium(III) chloride, indicating that the formation of the hydrogen peroxide complex of titanium(IV) increased upon the addition of titanium(III). We assumed that the copper(I) species produced by the reduction of titanium(III) induces the reduction of oxygen to produce hydrogen peroxide, as shown in the following equation:

  

$${\mathrm{O}}_2+2{\mathrm{H}}^{+}+2{\mathrm{e}}^{-}={\mathrm{H}}_2{\mathrm{O}}_2\hspace{1em}\hspace{1em}E°/\mathrm{m}\mathrm{V}=695$$(4) ¹¹⁾

Thus, when copper(II) is reduced by titanium(III) chloride, the formed chloro-complex of copper(I) acts as a catalyst, reducing the trace amount of oxygen contained in the nitrogen gas. Note that the profile of the redox potential was not affected by copper above −300 mV, at which iron reduction was complete, regardless of copper concentration.

3.3.2. Reduction Reaction of Vanadium

Vanadium exists in aqueous solutions as multiple chemical species with different valences. To investigate the relationship between the redox potentials of the solutions and the distribution of the vanadium species in the solution, the redox potential and absorption spectra of a hydrochloric acid solution containing only vanadium (V) without iron or tin were monitored as a function of the volume of the titanium (III) chloride aqueous solution. Figures 9(A) and 9(B) show the changes in the absorption spectra and redox potential and the absorbance of the vanadium solution at 766 nm, respectively. The absorption band at approximately 766 nm is assigned to tetravalent vanadium species, whereas the band at 399 nm reflects the absorption exhibited by pentavalent and trivalent vanadium species. When titanium (III) chloride was added to the vanadium solution, the redox potential began to decrease and reached the buffering zone in the range of 0.72 to 1.3 mol of titanium (III) chloride, as shown in Fig. 9(B). Meanwhile, the absorbance at 766 nm increased linearly with the amount of titanium(III) chloride added up to 0.72 mol and then decreased linearly, as shown in Fig. 9(B). These results indicate that vanadium(IV) evolved from vanadium(V) with the addition of up to 0.72 mol of titanium(III) chloride. When the amount of titanium(III) chloride added reached 0.72 mol, all vanadium(V) disappeared, and the potential dropped rapidly. The further addition of titanium (III) facilitated the reduction of vanadium (IV) to vanadium (III). A potential buffering zone for V(III)/(IV) was formed at approximately 100 mV. When the reduction of iron to divalent ions was complete, the vanadium species assumed the tetravalent state. Further addition of titanium(III) chloride reduced the tetravalent vanadium species to the trivalent state. The redox reactions involving vanadium (V), (IV), and (III) are expressed by the following equations:

  

$$\begin{array}{l}{\mathrm{V}}_2{\mathrm{O}}_5+6{\mathrm{H}}^{+}+4{\mathrm{e}}^{-}=2\mathrm{V}{{\mathrm{O}}_2}^{+}+3{\mathrm{H}}_2\mathrm{O}\\ E°/\mathrm{m}\mathrm{V}=0.958\end{array}$$(5) ¹¹⁾

  

$$\mathrm{V}{{\mathrm{O}}_2}^{+}+2{\mathrm{H}}^{+}+{\mathrm{e}}^{-}={\mathrm{V}}^{3+}+{\mathrm{H}}_2\mathrm{O}\hspace{1em}\hspace{1em}E°/\mathrm{m}\mathrm{V}=0.337$$(6) ¹¹⁾

The reduction of vanadium by titanium(III) chloride showed almost the same redox potential profiles under a nitrogen atmosphere, even when either vanadium(V) or vanadium(IV) was reduced. However, the profiles of the redox potential of the vanadium solutions with the addition of titanium(III) chloride under atmospheric conditions differed significantly from each other because of the difference in the original valence of vanadium. Figure 10 shows the changes in the redox potentials of the vanadium solutions upon the addition of potassium dichromate after reduction by titanium(III) chloride. In the case of vanadium(IV), the potential buffering zone of V(III)/V(IV) appeared at approximately 100 mV, as depicted in Fig. 10(A), whereas when vanadium (V) was used, the potential buffer zone appeared at approximately 400 mV. This suggests that different tetravalent and pentavalent vanadium species evolved individually during reduction by titanium(III) chloride.

Fig. 9. (A) Absorption spectra of V(V) solution after addition of TiCl₃. (B) Plots of absorbance (〇) at 766 nm and redox potential (□) of V(V) solution vs. amounts of TiCl₃ added. Experimental conditions: 70 mL of 0.3 mol/L hydrochloric acid solution containing 9.2×10⁻⁵ mol NH₄VO₃ was used to measure redox potential at room temperature under nitrogen atmosphere. Aliquots of a 0.13 mol/L TiCl₃ solution containing 6 mol/L HCl were added. Reference solution used was a 0.3 mol/L hydrochloric acid solution.

Fig. 10. Plots of absorbance (□) at 766 nm and redox potential (〇) of V(IV) solution (A) and of V(V) solution (B) vs. amounts TiCl₃ added. Experimental conditions: 70 mL of 0.3 mol/L hydrochloric acid solutions containing 9.2×10⁻⁵ mol NH₄VO₃ as V(V) and 9.2×10⁻⁵ mol VOSO₄ as V(IV) was used to measure redox potential at room temperature under nitrogen atmosphere. A 0.15 mol/L K₂Cr₂O₇ solution containing 6 mol/L HCl was added. A reference solution is a 0.3 mol/L hydrochloric acid solution.

Vanadium(V) exists in an acidic solution as a dioxyvanadium ion (VO₂⁺), in which two oxygen atoms bond to the central vanadium. When vanadium(V) is reduced, the evolved titanium(IV) is bound to the evolved vanadium(IV) via its two oxygen atoms, forming a hydrogen peroxide complex of titanium(IV).¹³⁾ The formed bi-nuclear complex provides a potential buffering zone at approximately 400 mV, which positively interferes with the subsequent dichromic acid titration. Strict control of the addition of titanium(III) chloride until the completion of the reduction of the iron species to the divalent state can completely suppress the formation of the bi-nuclear complex, such that the potential buffering zone at 400 mV is not observed. Thus, the stoichiometrically equivalent reduction of iron species enables the suppression of vanadium interference.

3.3.3. Reduction of Selenium

In this section, the roles of selenium(IV) and selenium(VI) in these redox profiles are investigated. Figure 11 shows the changes in the redox potential of the acidic iron solutions containing selenium(VI) or selenium(IV) as a function of the amount of added titanium (III) chloride. In the acidic iron solution containing selenium(VI), the addition of titanium(III) reduced selenium(VI) to selenium(IV) and provided a potential buffering zone for selenium(VI)/(IV) at 200 mV. Further addition of titanium(III) beyond the buffering zone reduced selenium(IV) to selenium(0) and produced red precipitates. In the case of iron solution with selenium(IV), the addition of titanium(III) yielded selenium(0), which directly reduced to form red precipitates. The potential buffering zone of selenium(VI)/(IV), which is almost the same as that of the copper-chloro complex, appears just below that of iron(II)/iron(III). Owing to the high redox potential of selenium(VI)/(IV), selenium(VI) may interfere with the subsequent titration of potassium dichromate and the chloro-complex of copper. This interference can be avoided by adding titanium(III) chloride solution when the reduction of iron is complete, as in the case of copper and vanadium. Again, the difference in the amount of titanium added at the drop point of the potential around 300 mV in Fig. 11 was due to the difference in the amount of tin(II) chloride added in advance, as shown in Fig. 7. The amount of titanium added at the drop point has no chemical meaning.

Fig. 11. Plots of redox potential of Fe(III) solution containing Se(VI) and Se(IV) vs. amounts of TiCl₃ added at 25°C under nitrogen atmosphere. Experimental conditions: 70 mL of 0.3 mol/L hydrochloric acid solution containing 4.6×10⁻³ mol Fe(III), 2.2×10⁻³ mol SnCl₂ and 4.6×10⁻⁵ mol K₂SeO₄ or 4.6×10⁻⁵ mol Na₂SeO₃ was used to measure redox potential at 25°C under nitrogen atmosphere. Aliquots of 0.13 mol/L TiCl₃ solution containing 6 mol/L HCl were added.

3.4. Strict Reduction of Iron by Titanium (III) Chloride

Based on the results described in Sections 3.2 and 3.3, the strict reduction of iron species in a digested solution of an iron ore can be achieved by controlling the redox potential between the completion of the reduction of iron(III) and the start of the reduction of tin(IV) and other diverse elements. Because these elements are not reduced at the redox potential, a strict reduction of iron can avoid any interference from these elements. Strict potential control can be achieved by adding titanium (III) chloride under a nitrogen atmosphere. This procedure is shown in Fig. 2 and is described in detail in Section 2.3.

Table 1 summarizes the results of the examination of the influence of diverse elements according to the protocol that combines the abovementioned reduction and potassium dichromate titration. The effects of the diverse elements were evaluated using relative values with respect to a hydrochloric acid based acidic iron aqueous solution containing no diverse elements, whose value was set to 100. None of the elements studied interfered with the titration, even when their concentration relative to iron was 1 mol%. We compared the proposed method with chelate titration using ethylenediaminetetraacetic acid, which is the most reliable method for determining metal ions. Because chelate titration with ethylenediaminetetraacetic acid lacks selectivity, appropriate separation is required to remove species that may cause potential interference in advance to determine the iron content in digested solutions of iron ores containing diverse metal elements. We prepared an acidic iron standard solution from iron chloride and compared the accuracy of the proposed method with that of chelate titration. The results of measurements repeated four times with both methods, which are summarized in Table 2, indicate that both analytical values agree with each other and that no significant difference was found in either the F test or the t test. Thus, the proposed method has precision and accuracy comparable to those of chelate titration using ethylenediaminetetraacetic acid.

Table 1. Effect of diverse ions on titrimetric determination of Fe(III). Experimental conditions: 70 mL of 0.3 mol/L hydrochloric acid solution containing 4.523×10⁻³ mol Fe(III), 2.2×10⁻³ mol SnCl₂ and 4.6×10⁻⁵ mol diverse metals was used to measure redox potential at 25°C under nitrogen atmosphere. Diverse elements were added as Co(NO₃)₂, Ni(NO₃)₂, ZnCl₂, Cu(NO₃)₂, NH₄VO₃, mol MnCl₂, K₂SeO₄ and Na₂SeO₃. A 1.667×10⁻² mol/L K₂Cr₂O₇ solution was added as a titrant.

	Diverse metal
	None	Ni (II)	Co (II)	Zn (II)	V (V)	Mn (II)	Se (VI)	Se (VI)	Cu (II)
Recovery of Fe (III)	1.000	1.000	0.999	1.000	1.000	1.000	1.000	1.000	1.000

Table 2. Comparison of the proposed method with chelating titration. Experimental conditions: 70 mL of 0.3 mol/L hydrochloric acid solutions containing 4.523×10⁻³ mol Fe(III) were titrated with 1.667×10⁻² mol/L K₂Cr₂O₇ solution potentiometrically and with 1.000×10⁻² mol/L EDTA·2Na solution complexiometrically.

Titration	Concentration of iron ion (n=4)/10−2 mol·L−1	F	t
K2Cr2O7	6.461±0.001	0.8816	0.5476
EDTA·2Na	6.461±0.001

Finally, certified reference materials for the iron ores were analyzed using the method described in this paper. As summarized in Table 3, the analytical values of all certified reference materials agreed well with the certified values, with no significant differences observed between the analytical and certified values.

Table 3. Determination of iron contents in certified reference materials. A 1.667×10⁻² mol/L K₂Cr₂O₇ solution used as a titrant for potentiometric titration.

CRM	Analytical results/wt%	Centified value/wt%	F	t
JSS 831-2 (Taharoa iron sand)	56.61±0.02 (n=4)	56.64±0.09 (n=10)	17.10	0.6978
JSS 850-3 (Marcona pellet)	66.77±0.02 (n=4)	66.78±0.05 (n=11)	4.257	0.6309
JSS 814-1 (Peru magnetite)	56.68±0.05 (n=4)	56.70±0.09 (n=11)	3.251	0.5185

4. Conclusion

This study investigated the redox reactions of iron species in digested solutions of iron ores using potentiometry and spectrophotometry under nitrogen and air atmospheres. Iron(III) was reduced by tin(II) chloride, followed by excess titanium(III) chloride. The remaining excess titanium(III) chloride was treated with potassium dichromate to achieve a strict reduction of the iron species to its divalent state. The following conclusions were drawn:

(1) The redox reaction of tin(II)/(IV) is extremely slow, significantly affecting the determination of the endpoints of reduction using indigo carmine as an indicator.

(2) The potential at which copper chloro-complex undergoes a redox reaction is higher than the potential at which the indicator changes color. The reduced Cu (I) species produces a positive error in the subsequent potassium dichromate titration.

(3) Pentavalent vanadium forms a complex with titanium as a result of a redox reaction with titanium(III) chloride, which positively interferes with the potassium dichromate titration.

The interference caused by the above elements can be avoided by strictly controlling the amount of titanium(III) chloride added to the digested solution of the iron ore to reduce only the iron species in the solution to its divalent state. To add titanium(III) chloride precisely and accurately, the redox potential of the digested solution was monitored under a nitrogen atmosphere. The iron content in certified standard materials of iron ores was accurately determined by reducing iron to a divalent state using the proposed method, followed by potassium dichromate titration.

References

• 1)  JIS M 8212: 2014, Iron ores - Determination of total iron content (in Japanese).
• 2)  ISO 9507: 1990, Iron ores - Determination of total iron content - Titanium(III) chloride reduction methods.
• 3)   M.  Saeki,  K.  Nishizaka,  M.  Iwamoto and  K.  Adachi: Tetsu-to-Hagané, 60 (1974), 2045 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.60.13_2045
• 4)  Steel Chemical Analyses Vol. 9, ed. by Japan Society for the Promotion of Science, No. 19th Committee, Nikkan Kougyo Shinbunsha, Tokyo, (1963), 28 (in Japanese).
• 5)  ISO 9508: 1990, Iron ores – Determination of total iron content – Silver reduction titrimetric method.
• 6)   W.  Scott: J. Ind. Eng. Chem., 11 (1919), 1135.
• 7)   E. R.  Riegel and  R. D.  Schwartz: Anal. Chem., 24 (1952), 1803. https://doi.org/10.1021/ac60071a025
• 8)   K.  Wakino,  M.  Murata and  M.  Omatsu: Bunseki Kagaku, 20 (1971), 395 (in Japanese). https://doi.org/10.2116/bunsekikagaku.20.395
• 9)   T.  Hata,  T.  Hagiwara and  K.  Sumi: Bunseki Kagaku, 22 (1973), 886 (in Japanese). https://doi.org/10.2116/bunsekikagaku.22.886
• 10)   H.  Hu,  Y.  Tang,  H.  Ying,  M.  Wang,  P.  Wan and  X. J.  Yang: Talanta, 125 (2014), 425. https://doi.org/10.1016/j.talanta.2014.03.008
• 11)  Handbook of Chemistry, 4th ed, Vol. II, ed. by The Chemical Society of Japan, Maruzen, Tokyo, (1980), 466 (in Japanese).
• 12)   C.  Nila and  I.  González: Hydrometallugy, 42 (1996), 63.
• 13)   M.  Ramachandran,  P.  Sami,  T.  Jeyabalan and  A.  Shunmugasundaram: Trans. Metal Chem., 23 (1998), 583.