Jiawei Wang,Dehua Che,Wei Qin*

State Key Laboratory of Chemical Engineering,Department of Chemical Engineering,Tsinghua University,Beijing 100084,China

Keywords:Solvent extraction Selectivity Rubidium t-BAMBP Cyclohexane

ABSTRACT 4-Tert-butyl-2-(α-methylbenzyl)phenol(t-BAMBP)was used in cyclohexane in the extraction ofrubidium from brine sources containing lithium.The effect of t-BAMBP concentration and aqueous phase pH on the rubidium and lithium extraction equilibrium was studied.t-BAMBP/cyclohexane was efficient and selective for rubidium extraction with optimal operating conditions being pH of 13.0 and initial t-BAMBP concentration of 1.0 mol·L-1.The stoichiometry of the complex of t-BAMBP with rubidium is 4:1.The apparent extraction equilibrium constant of rubidium was calculated by fitting the experimental data.

1.Introduction

Rubidium,a metallic element with active chemical properties,has the second lowest ionization potential of the non-radioactive isotope alkali metals after cesium.Because of its unique qualities,rubidium is used widely in national defense,aerospace industry,genetic engineering,medicine,energy industry,and environmental science[1,2].

Although it is one of the more abundant elements in the Earth crust,rubidium occurs only at low concentration in rubidium minerals and rubidium-rich waters,along with otheralkalimetals[3,4].Because alkali metals do not easily form precipitates or complex ions and have similar aqueous chemistry,the separation and purification of rubidium is difficult.

Solvent extraction is an economical,efficient,and environmentally friendly method for the separation of dilute metal ion solutions.In 1962,Brown[5–8]used substituted phenols instead of amines for the solvent extraction of137Cs from fission product wastes.Later investigations[9–13]found that the order of extractability by 4-sec-butyl-2-(αmethylbenzyl)phenol(s-BAMBP)was Cs＞Rb＞K＞Na＞Li,in a strongly alkaline environment.The separation coefficient for rubidium/potassium was also high and substituted phenols would be good extractants for the separation of rubidium.4-Tert-butyl-2-(α-methylbenzyl)phenol(t-BAMBP)has a similar extraction behaviorto s-BAMBP,but is less difficult to synthesize[14–16].t-BAMBP is also considered to be a favorable extractant for rubidium separation[17,18].

In China,most of the available rubidium resource dissolves in the salt lake.In a usual recovering process,rubidium ions generally coexist with lithium and cesium ions in the salt lake brine after pre-treatment.In order to recover the rubidium efficiently and reduce the loss of lithium simultaneously,the extraction behavior of rubidium and lithium by t-BAMBP is investigated in this work.The extraction mechanism is studied and the apparent extraction equilibrium constant is calculated.Results obtained would be helpful for future study and in process design for recovery of rubidium.

2.Materials and Methods

2.1.Chemicals

Inorganic reagents,RbCl(Shanghai Dibai Chemical Reagent Tech Co.,purity＞99.5%)and LiCl(Beijing Yili Fine Chemical Co.,purity＞97%)and organic reagents,t-BAMBP(Beijing Realkan Seprtech Co.,purity＞95%)and cyclohexane(Beijing Modern Eastern Fine Chemical Co.,purity＞99%),were used in the present work.

2.2.Experimental

All extraction experiments were conducted in 50-ml flasks at(25 ± 2)°C.The simulated salt lake brine sample was prepared consulting the practical pre-treated salt lake brine,containing 200×10-6Rb+and 200×10-6Li+.The pH of the aqueous solution sample was adjusted from 11 to 14 by addition of an alkaline solution.Aqueous and organic solutions(10 ml each;0.2,0.4,0.6,0.8,and 1.0 mol·L-1t-BAMBP in cyclohexane)were added to the flasks,which were shaken manually for 10 min and left to stand for 30 min for phase disengagement.Preliminary experiments showed that equilibrium was obtained after approximately 2 min.

2.3.Analysis

Aqueous sample pH was determined using a PHS-25C pH meter(Mettler,Switzerland).Rubidium and lithium concentrations were determined using a VISTA-MPX inductively coupled plasma-atomic emission spectrometer(Varian,USA)and Z-5000 atomic absorption spectrometer(Hitachi,Japan).Organic phase rubidium and lithium concentrations were calculated by material balance.Preliminary experiments indicated that the deviation in concentration value was±3%.

To investigate intermolecular hydrogen bonding,infrared analysis was carried out using a Bruker Tensor 27 Fourier transform infrared spectrometer and OPUS data collection program,by injecting initial organic solution into a sample cell with constant thickness.Spectra were recorded at 1 cm-1resolution between 4000 and 3000 cm-1and averaged over 32 scans.All spectra were subtracted to remove the influence of diluent absorption and normalized using the aromatic C–H stretching vibration peak at 3028 cm-1to remove the influence of concentration.

2.4.Parameter definition

The distribution coefficient is defined as D=Co,eq./Ca.eq.,where Co,eq.and Ca.eq.are metal ion concentrations in the organic and aqueous phases,respectively.The separation coefficient is defined as βRb/Li(=DRb/DLi).

3.Results and Discussion

3.1.Association of t-BAMBP molecule

t-BAMBP is a substituted phenol.The infrared spectra in Fig.1 at different concentrations of t-BAMBP in cyclohexane are typical of those observed for phenols that undergo intermolecular hydrogen bonding.The infrared analysis of organic phase was accomplished in a sample cell with constant thickness,and the spectra in different concentrations were normalized in quantity.A sharp band characteristic of non-associated O–H in the monomer is observed at 3618 cm-1and a broader characteristic of the hydrogen-bonded O–H in the dimer is observed at 3547 cm-1.An absorption region is observed at 3200–3500 cm-1,where the O–H stretching vibration characteristics of more highly associated,hydrogen-bonded species are expected.As the t-BAMBP concentration increases,the proportion of monomer decreases and the proportion of more highly associated species increases.In this process,the proportion of dimer remains almost unchanged.

Fig.1.Infrared spectra of initial organic solutions of various t-BAMBP concentrations(1–0.2 mol·L-1;2–0.4 mol·L-1;3–0.6 mol·L-1;4–0.8 mol·L-1;5–1.0 mol·L-1).

From the extraction characteristics and mechanism analyses[12,13,16,20],the dimer is found to be the predominant substituted phenol under the experimental conditions.The extraction reaction is described as a cation exchange reaction,which can be proposed as follows:?

where M and ROH are the metallic element and substituted phenols(t-BAMBP),respectively and subscripts a and o denote aqueous and organic phases,respectively.

3.2.Effect of pH and t-BAMBP concentration

According to Eq.(1),the extraction equilibrium will be affected by aqueous phase pH.The experimental results are presented in Figs.2–4,including distribution coefficients,DRband DLi,and separation coefficient,βRb/Li,for different aqueous pH values and t-BAMBP concentrations.Both DRband DLiincrease with pH,and DLiincreases more sharply a thigh pH than at low value.A peak value ofβRb/Liexists around pH 13.

The p Kavalue for Li+in aqueous solution is 13.82[19],which is much lower than that for Rb+.Consequently,the rubidium ions maintain the steady-state in the alkaline solution,while the lithium ions tend to combine with hydroxyl ions.The study of hydration indicates that lithiumions exist as Li(H2O)4+in aqueous solution,forming a tetrahedron[21].Therefore,the transmutation is as follows

LiOHis more likely to enter the organic phase than Li(H2O)4+according to their polarity,which leads to a rapid ascension in the distribution coefficient of lithium with high pH.

The t-BAMBP concentration has an appreciable influence on extraction behavior,as shown in Figs.2–4.The distribution coefficient of rubidium and the separation coefficient of rubidium/lithium are greater at higher t-BAMBP concentrations.Rubidium ions are therefore extracted into the organic phase more effectively and selectively.In the experimental range,rubidium recovery can exceed 90%and βRb/Liapproaches 100.

In practice,however,the organic phase viscosity is so high that phase disengagement becomes very difficult at concentrations above 1.0 mol·L-1.Therefore,t-BAMBP concentrations higher than 1.0 mol·L-1are not recommended.

Fig.2.D Rb vs.pH for various t-BAMBP concentrations(■ 0.2 mol·L-1;● 0.4 mol·L-1;▲0.6 mol·L-1;? 0.8 mol·L-1;★ 1.0 mol·L-1).

Fig.3.D Li vs.pH for various t-BAMBP concentrations(■ 0.2 mol·L-1;● 0.4 mol·L-1;▲ 0.6 mol·L-1;?0.8 mol·L-1;★ 1.0 mol·L-1).

Fig.4.βRb/Li vs.pH for various t-BAMBP concentrations(■ 0.2 mol·L-1;● 0.4 mol·L-1;▲ 0.6 mol·L-1;?0.8 mol·L-1;★ 1.0 mol·L-1).

3.3.Description of extraction equilibrium of rubidium

As shown in Eq.(1),complex(MOR)?(ROH)2n-1forms and is extracted rapidly into the organic phase.The apparent extraction equilibrium constant,K,is defined as:

and the distribution coefficient can be expressed as

Thus,

and its logarithmic form can be written as

In these experiments,the concentration of extracting reagent is two orders of magnitude higher than the concentration of rubidium and lithium.Therefore,it can be assumed that the t-BAMBP concentration remains constant during the extraction process.Moreover,K can be considered to be a constant value because of the low ion concentration.Therefore,a plot of lg Dvs.pH would be a straight line with a unity slope.Regression curves are shown in Fig.5 with equations given in Table 1.All slopes are approximately equal to 1.The assumption that the reaction is a cation exchange extraction is therefore feasible.

Fig.5.Relationship of lg D Rb and pH for various t-BAMBP concentrations(■ 0.2 mol·L-1;● 0.4 mol·L-1;▲ 0.6 mol·L-1;?0.8 mol·L-1;★ 1.0 mol·L-1).

Table 1 Regression results for lg D Rb vs.pH for various t-BAMBP concentrations

According to Eq.(6),a plot of(lg DRb–pH)vs.lg[(ROH)2]yields a straight line with a slope of n and an intercept lg KRb,as shown in Fig.6.The slope of the fitting line is approximately equal to 2.Therefore,n=2 in Eq.(1).The reaction equation can be written as

The experimental data are fitted using y=2x+C to calculate KRb.From the curve intercept,the apparent extraction equilibrium constant,KRb,is found to be 3.44 × 10-12L·mol-1.The calculated extraction equilibrium is obtained from the fitting parameters.The calculated and experimental distribution coefficients of rubidium(Dcaland Dexp,respectively)are compared in Fig.7 with satisfactory results.

The(MOR)?(ROH)3complex is in accordance with a plausible structure proposed by Egan et al.[13],as shown in Fig.8.This suggested structure maintains solvation by hydrogen bonds between pairs of solvating phenols.The stoichiometry of 4:1 is also provided by Chen and Chen[17],from saturated extraction experiments.

Fig.7.Comparison of D cal and D exp.

Fig.8.Possible structure of the complex.

4.Conclusions

The extraction and separation behavior of rubidium by t-BAMBP/cyclohexane were investigated.It is found that t-BAMBP is an efficient and selective extractant for rubidium.t-BAMBP concentration and aqueous pH influence the extraction equilibrium significantly.The resultant complex is considered to be(MOR)?(ROH)3.The apparent extraction equilibrium constant of rubidium is 3.44 × 10-12L·mol-1at 25°C.

Nomenclature

C fitting curve intercept

Ca.eq.equilibrium concentration in the aqueous phase

Co,eq.equilibrium concentration in the organic phase

D distribution coefficient

Dcalcalculated distribution coefficient

Dexpexperimental distribution coefficient

DLidistribution coefficient of lithium

DRbdistribution coefficient of rubidium

K apparent extraction equilibrium constant

KRbapparent extraction equilibrium constant of rubidium

n stoichiometric factor

βRb/Liseparation coefficient of rubidium/lithium