Solvent Extraction of Fe ( III ) from Aqueous Chloride Solution by Cyanex 301 Dissolved in Kerosene

The title system has been investigated over a wide range of aqueous acidity. The equilibration time is 1 h. The extraction ratio (D) is independent of [Fe(III)] provided equilibrium [HCl] and [HA] are kept constant. At a constant equilibrium extractant concentration, the [HCl] dependences are -1.6, ~0 and -3 in the [HCl] regions of >3, 2-0.5 and <0.3 M; respectively; whilst at constant [HCl], the [HA] dependence is 3.0. On the other hand, [Cl] dependence varies within -0.5 to -3 at constant [HCl] of 0.3 M; whereas its values are ~ -1 and ~ 0.63 at constant [HCl] of 3 and 1 M, respectively. Based on these results the extraction mechanisms have been suggested to be + 3 HA(o) ⇌ FeA3(o) + n Cl + 3 H in the low [HCl] region, ( ) + − n n 3 FeCl + 3 HA(o) + n Cl ⇌ FeCl3.3HA(o) in the intermediate [HCl] region and HFeCl4 + 3 HA(o) ⇌ FeCl3.3HA(o) + HCl in the high [HCl] region under investigation. The Kex and ΔH values have been evaluated. Loading capacity is 5.5 g Fe(III)/100 g Cyanex 301. The stripping can be made effective by a mixture of 6 M H2SO4 and 1 M Na2C2O4.


Introduction
The green viscous bis-(2,4,4-trimethylpentyl)dithiophosphinic acid (trade name being Cyanex 301, C 16 H 34 PS 2 H, HA) has been introduced as an extractant in the last decade of the 20 th century by American Cyanamide Co. and Cytec Canada Inc. Supplied sample contains 77.2% bis (2,4,4-trimethylpentyl)dithiophosphinic acid and there is still no way to purify it more.Some physical constants of the supplied Cyanex 301 are given elsewhere [1].The vapor pressure osmometry (VPO) analysis indicates Cyanex 301 is monomeric [2].
Iron-bearing compounds are often present as gangue materials in ores of many common valuable metals.Therefore, it becomes necessary to separate iron from a leached solution of an ore for the production of a pure metal hydrometallurgically and this can be carried out effectively by the solvent extraction technique.

Reagents
Cyanex 301 was donated by Cytec Canada Inc. Kerosene was obtained from the local market and distilled to collect the fraction distilling over 200-260 o C. It was colorless and mostly aliphatic in nature.Ferric chloride (Loba Chemie, 99%) was used as a source of Fe(III).All the other chemicals were of reagent grade and used as received.

Analytical
The concentration of Fe(III) in the aqueous phase was determined by the thiocyanate method [29] at 480 nm using a WPA S104 Spectrophotometer.The standard solution of Fe(III) was prepared by dissolving 0.846 g A. R. FeNH 4 (SO 4 ) 2 in 1 L 0.1 M H 2 SO 4 solution (1 mL = 0.1 mg Fe(III)).The acidity of the aqueous phase was adjusted by the addition of HCl; whilst, Cl -concentration by NaCl addition.

Extraction procedure
A stock solution of FeCl 3 (1 L) was prepared to contain 10.01 g Fe(III) (0.179 M), 0.12 M H + and 0.657 M Cl -.This solution was used to prepare the aqueous phases containing different amounts of H + , Cl -and Fe(III) for extraction.The extraction procedures are given elsewhere [4,12,15].Equal aliquots of organic and aqueous phases (20 mL each) were taken in an 125 mL reagent bottle and agitated for a predetermined time (1 h) at 303 K in a thermostatic water bath.After mechanical shaking, the phases were allowed to settle, separated and the aqueous phase was analyzed for its Fe(III) content colorimetrically as mentioned above.The concentration of Fe(III) in the organic phase was calculated from the difference.The value of the distribution or extraction ratio (D) was calculated as the ratio of the concentration of Fe(III) in the organic phase to that existing in the aqueous phase at equilibrium.

Loading procedure
Loadings of Fe(III) in 0.177, 0.147 and 0.118 M Cyanex 301 solutions were carried out by vigorous contact of these phases (50 mL) separately and repeatedly with fresh aqueous solutions (containing 1 g/L Fe(III); and 0.3, 1.0 and 3.0 M HCl for 0.177, 0.147 and 0.118 M extractant systems, respectively) of same volume until the organic phases were saturated with Fe(III).After each contact, the phases were disengaged and the aqueous phases were analyzed for their Fe(III) contents.The amount of Fe 3+ transferred into the organic phase for each contact was calculated from the difference and then the cumulative concentrations of Fe(III) in the organic phase (cumulative [Fe(III)] (o) , g/L ) after each stage of contact were determined.

Stripping procedure
The loaded organic phases obtained above were diluted separately with kerosene so that the resultant solutions contained 1 g/L Fe(III) as complex and practically no free extractant.These solutions were used to study stripping of various Fe(III)-HA entities existing in the organic phase depending on aqueous acidities used in extractions.Strippings were performed by 1 M and 6 M H 2 SO 4 , HNO 3 , HCl and HClO 4 acid solutions together with a mixture of 6 M H 2 SO 4 and 1 M Na 2 C 2 O 4 .In stripping, 10 mL of Fe(III) loaded organic phase was equilibrated with an equal aliquot of each of the above acid solution for 1 h at 303 K.After equilibration, the phases were settled, disengaged and the aqueous phase was analyzed for Fe(III) content.In stage-wise stripping, the organic phase was recycled with equal aliquot of fresh aqueous phase.

Results and Discussion
Preliminary experiments show that the concentration ratio ([Fe(III)] (o) /[Fe(III)] (aq) ) increases almost exponentially with increasing phase contact time up to 45 min for the investigated system which indicates that the equilibration time for the system is 45 min.In subsequent experiments, the phase contact time of 1 h was allowed to ensure equilibration under different experimental conditions.Previously, it has been reported that equilibration times for the extraction of Fe(III) from chloride medium by analytical grade D2EHPA [12], technical grade D2EHPA [15], Cyanex 272 [20] and Cyanex 302 [30] in kerosene are 50 min.So the equilibration time for the extraction of Fe(III) from chloride medium by Cyanex 301 in kerosene is similar to those by other acidic organo-phosphoric and phosphinic acid derivatives in kerosene.Fig. 1 shows the variations of D with initial [Fe(III)] (3.72 mM -64.87 mM) while extracted by 0.059 M Cyanex 301 at 0.3, 1 and 5 M HCl in log-log scale.It is found that in all cases the distribution ratio is decreased extensively with the increase of initial Fe(III) concentration in the aqueous phase at the cited constant initial concentrations of HCl and HA; particularly at higher concentration region of Fe(III).This type of behavior indicates the formation of non-extractable Fe(III)-Cl -species in the aqueous phase or the scarcity of extractant at higher Fe(III) concentration.However, this statement will be valid only when the equilibrium acidity and extractant concentration remained constant [31].Since relatively high concentration of HCl has been used in the study, there will be little change in its concentration on equilibration.In contrast, the equilibrium concentration of free HA will be decreased from the initial concentration to a greater extent on extracting large amounts of Fe(III) particularly from concentrated Fe(III) solutions.On considering the extracted species being 1 : 3 Fe(III) -HA complexes (as will be seen latter), the necessary corrections of log D values to get log C D values have been made (where, C D is the corrected D values at constant equilibrium HCl and HA concentration equaling to constant initial HCl and HA concentrations) as follows: log  The plots are straight lines with almost zero slopes (the least squares slopes are -0.068,0.058 and 0.085 for 0.3, 1.0 and 5.0 M HCl systems, respectively).It is therefore concluded that the extraction ratio is independent of Fe(III) concentration in the aqueous phase provided the equilibrium [HCl] and [HA] are kept constant; and this behavior is in consistent with the principle of solvent extraction.Results also suggest that the percentages of various Fe(III)-Cl -species are not changed with Fe(III) concentration provided sufficient Cl -exists there, which is in accord with Gamlen and Jordan [32].They suggested that the percentages of various Fe(III) -Cl -species existing in the aqueous phase were a function of Cl -concentration in the system, but not of [Fe(III)].It is concluded from these results that the extractant dependence remains unaltered over the entire HCl concentration range used in this study and the extractant concentration functionality is very high for this system.Whatever may be the [HCl] in the system, the extractant dependence is always 3; i.e., 3 moles of extractant are attached to 1 g ion of Fe(III) to form 1 mole of extractable species., the extraction ratio is decreased with increasing [Cl -], but the extent of decrement depends on constant [H + ] used in the system and the [Cl -] region as well.From this result, it can be concluded that at least on a qualitative basis Cl -is liberated by the extraction reaction when the [HCl] is kept either high or low.On the other hand, at intermediate [H + ] region (~ 1.0 M), the extraction ratio is increased with increasing [Cl ]; so that Cl -is added to the existing Fe(III)-Cl -species in the aqueous phase to form the extractable species.The literature reports [12,32] indicate that the existing Fe(III) species in 0.2 M Cl - medium are 12.9 % Fe 3+ , 38.1 % FeCl 2+ , 35.6 % FeCl 2 + and 13.4% FeCl 3 ; whereas in 0.4 M Cl -medium, the respective percentages are changed to 7.5, 27, 42.  2) can explain the experimental data obtained around 1 M HCl medium.
The ion pair, HFeCl 4 starts to form in the 3 M Cl -medium [12,32] and so in the higher [HCl] region, the extraction occurs via the following reaction: Eq. (3) satisfies the extractant dependence of 3 but it indicates the HCl-dependence of -1.However, experimentally [HCl]-dependence of about -1.6 is obtained.This might be due to the presence of H x FeCl 3+x (x > 1) in the medium at still higher [HCl] region.From the foregoing discussion, it is revealed that the extracted species is FeA 3 at the lower [HCl] region and that at intermediate and higher  3) are exothermic in nature and its high value at 0.3 M HCl system supports the formation of extractable species by chelation; whereas, low values at intermediate and higher [HCl] levels support the formation of extractable species by solvation.The loading capacity, defined as the maximum amount of metal ion in gram extracted per 100 g of a pure extractant, is a very important factor for an extractant's commercial applicability.High loading capacity is desirable for a particular extractant-metal ion system.Moreover, the species extracted at high loading may be easily converted to pure (usually solid) complexes for its structure determination by chemical and instrumental analyses.The cumulative [Fe(III)] (o) vs. contact number plots are given in Fig. 6 (when the organic phases were repeatedly contracted with fresh aqueous phases) at three different sets of experimental parameters.It is indicated that most of the Fe(III) existing in the aqueous phase is extracted into the organic phase up to the 4 th contact and then the uptake by Cyanex 301 is gradually decreased to zero at the 8 th contact in all three cases.It is found from Fig. 6 [12], 13.13 g Fe(III)/ 100 g technical grade D2EHPA [15], 9.6 g Fe(III)/100 g Cyanex 272 [20] and 29.41 g Fe(III)/ 100 g Cyanex 302 [30].Moreover, the maximum loading capacity suggests that the extracted complex should have a Cyanex 301/Fe(III) ratio of 3 in the complex (FeA 3 or FeCl 3 .3HA),which has already been established by the mechanistic study.So the extraction mechanism at a certain aqueous acidity is not changed with the extent of loading.From the intercepts of the lines in Figs. 2 and 3, the value of extraction equilibrium constants (K ex ) for the extraction of Fe(III) from three distinct regions of [HCl] have been evaluated and tabulated (Table 1  a) log Kex = I -3 log [Cyanex 301]eq (b) log Kex = I + x log [HCl](eq); where x = 3, 0 and 1.6 in the top, middle and bottom cases, respectively.
The strippings of Fe(III) from the organic phase (extracted at various aqueous acidities) by various acid solutions have been investigated and the results have been tabulated (Table 2).It is seen that 1 or 6 M mineral acids alone (H 2 SO 4 / HNO 3 /HClO 4 /HCl) are not good stripping agents for bringing back Fe(III) from the extracted complexes formed at various aqueous acidities even when the organic phases do not contain practically no free extractant.However, the mixed stripping agent consisting of 6 M H 2 SO 4 and 1 M Na 2 C 2 O 4 is found to be effective for stripping, which can strip ~ 60 % Fe(III) in single stage and cumulative ~ 99% Fe(III) in the 4 th stage of stripping.It is also noticed that stripping of Fe(III) from FeA 3 (formed at [HCl] = 0.3 M) is little more difficult than from FeCl 3 .3HA(formed at [HCl] = 1 and 3 M).It is therefore recommended to use 6 M H 2 SO 4 and 1 M Na 2 C 2 O 4 mixture as stripping agent.

Conclusion
The following conclusions can be drawn:
Fig.1shows the variations of D with initial [Fe(III)] (3.72 mM -64.87 mM) while extracted by 0.059 M Cyanex 301 at 0.3, 1 and 5 M HCl in log-log scale.It is found that in all cases the distribution ratio is decreased extensively with the increase of initial Fe(III) concentration in the aqueous phase at the cited constant initial concentrations of HCl and HA; particularly at higher concentration region of Fe(III).This type of behavior indicates the formation of non-extractable Fe(III)-Cl -species in the aqueous phase or the scarcity of extractant at higher Fe(III) concentration.However, this statement will be valid only when the equilibrium acidity and extractant concentration remained constant[31].Since relatively high concentration of HCl has been used in the study, there will be little change in its concentration on equilibration.In contrast, the equilibrium concentration of free HA will be decreased from the initial concentration to a greater extent on extracting large amounts of Fe(III) particularly from concentrated Fe(III) solutions.On considering the extracted species being 1 : 3 Fe(III) -HA complexes (as will be seen latter), the necessary corrections of log D values to get log C D values have been made (where, C D is the corrected D values at constant equilibrium HCl and HA concentration equaling to constant initial HCl and HA concentrations) as follows: log C D = log D + 3 [log [HA] (o,ini) -log ([HA] (o,ini) -3 [Fe(III)] (o,eq) )] -x [log [HCl] (ini) -log [HCl] (ini) -x [Fe(III)] (o,eq) ]; whence x is [HCl] dependence and it depends on [HCl] region: x = -3, 0 and -1.6 at low, intermediate and high concentration regions of HCl, respectively.The log C D vs. log [Fe(III)] (ini) plots are also given in Fig. 1.

Fig. 4
Fig. 4 shows the dependence of extraction ratio on chloride ion concentration as log
i) Cyanex 301 can effectively extract Fe(III) from chloride solution.The equilibration time is 1 h.ii) The extent of extraction is independent of the initial Fe(III) concentration in the aqueous phase provided equilibrium aqueous acidity and Cyanex 301 concentration are kept constant.iii) The breaks in [HCl] dependence curve indicate three types of extraction reactions depending on the [HCl] prevailing in the system.The [HCl] dependences are -3, ~ 0 and -1.6 in its lower, intermediate and higher concentration regions, respectively.The extractant dependence is always 3. iv) Both [HCl] and [HA] dependencies of 3 but of opposite sign suggest that the extraction occurs via the reaction: Fe 3+ + 3HA (o) ⇌ FeA 3 (o) + 3 H + with the value of extraction equilibrium constant (K ex ) of 10 3.064 in the low [HCl] region.v) At intermediate [HCl] region, the [HCl] dependence of ~ 0 and [HA] dependence of 3 leads to the conclusion that the extraction under this condition occurs via the reaction: FeCl 3 + 3 HA (o) ⇌ FeCl 3 .3HA(o) with K ex value of 10 3.986 .vi) At higher [HCl] region, the extraction proceeds via reaction: HFeCl 4 + 3HA (o) ⇌ FeCl 3 .3HA(o) + HCl with K ex value of 10 4.621 .vii) The extraction is found to depend on Cl -concentration because of changes in the % composition of Fe(III)-Cl -species in the aqueous phase.

Table 1 .
[30]e average values of log K ex are 3.064, 3.986 and 4.621 with respective standard deviations (of log K ex ) of 0.136, 0.067 and 0.104 for extractions of Fe(III) from lower, intermediate and higher HCl concentration regions, The value of log K ex for Fe(III)-Cyanex 272 and Fe(III)-Cyanex 302 systems are -2.3[20]and-0.632[30],respectively.Although the extraction equilibrium constant at any [HCl] region is many times higher in case of Cyanex 301 than in cases of Cyanex 272 and Cyanex 302, the high extractant functionality or dependence in the present case at both high and low loadings renders low loading capacity of Cyanex 301 towards Fe(III).Consequently, Cyanex 301 comes out as a less effective commercial extractant for Fe(III).Elucidation of extraction equilibrium constants (K ex ) at various aqueous acidity regions for the extraction of Fe(III) by Cyanex 301 at 303 K.

Table 2 .
Data for stripping of extracted Fe(III)-Cyanex 301 complex from kerosene phase by various acid solution.