ION-EXCHANGE STUDIES WITH SOME TITANIUM COMPOUNDS AND THEIR MEMBRANES

Abstract

Although quite a good deal of literature is available on inorganic ion~exchange gels, certain class of compounds are yet to be completely explored. Simple and mixed salts of arsenate anion have either not been tried or the data available on these compounds are quite scanty. These inorganic gels also provide a highly suitable material for the development of ion-exchange membranes. As such investi gations were planned on arsenate gels and the preparation of titanium arsenate and tungstoarsenate and their exchange as well aS membrane properties formulate the subject matter of this thesis. Titanium tungs toarsenate is prepared by mixing tita nium(IV) chloride (0.25 M) to a mixture of sodium arsenate and sodium tungstate (0.25 M each, mixing ratio 2:1:1) at pH one. This compound is white in colour and is stable in acids and salt solutions but dissolves in bases especially sodium hydroxide. The stability of this exchanger material in acids and salt solutions is higher as compared to titanium arsenate. The ion-exchange capacity of the product is found to be 0.86 meq per g and decreases when dried at higher temperatures. It is interesting to note that the equilibra tion time for this exchanger is much less (10 h) than what has normally been reported for other synthetic inorganic ion exchangers. The product has been characterized on the ii basis of chemical analysis, I.R. spectra, the rmogravimetric analysis and X-ray diffraction results. The compound is amorphous in nature, the ratio of Ti: W: As is 1: 4; 2 and a tentative formula Tio2( W03)4( As^) .16^0 is assigned to it. On ths basis of distribution coefficient (K,) values of various metal ions titanium tungs to arsenate is found to be selective for Rb+. Besides this, the material also exhibits a significant uptake of Cs+, Tl+, Cd2+, Hg2+, Sr2+, Ba2+, 2+ 3+ Zn and Pe^ whereas the sorption of other cations are almost negligible. The uptake of metal ions decrease with increasing NH^HOj or HNO^ concentration. The slope of log K, vs log NH^NO^ or HNO, concentration plots is less than the valency of exchanging ion thereby indicating that the sorption of ions do not proceed strictly through ion-exchange mechanism. On the basis of selectivity pattern of this exchanger material, a large number of binary separations have been performed on titanium tungstoarsenate column. The separations are quite clean, without any contamination and recovery is hundred percent. The compound is found to be a good scavenger of Rb+ ions. It is also worth mentioning that except Rb+, Cs+, Tl+ and Ag+, all the metal ions could be eluted with water at two pH. Adsorption of some uni-, bi- and trivalent metal ions (pH 5-6) on titanium arsenate and titanium tungs to arsenate gels has also been investigated. It is observed that except Ill univalent ions, the adsorption of metal ions become negli gibly small at pH 1.8. This is attributed to the competition offered by H+ ions in solution. The adsorption sequence of metal ions on the two gels are; Cs+>Tl+>Eb+ >Ag+, Pb2+ = Ba2+>Sr2+>Mn2+>Zn2+> Co2+>Mg2+>Ca2+i Or5+>Pe5+>Al5+ on titanium arsenate and Rb+>Cs+>Tl+>Ag+; Ba2+>Sr2+>Mn2+>Zn2+>Ca2+>Mg2+> Cd2+; Pe5+>Cr5+>Al5+ on titanium tungstoarsenate. This order is in accordance with increasing ionic potentials of the metal ions, with the exception of Tl+, Ca +, Co and Cr^+ in the case of titanium arsenate and Rb , Ca and Cd in the caSe of titanium tungstoarsenate. It is further observed that the maximum uptake (^gx.) value for Cs ion on titanium arsenate and that of Rb+ ion on titanium tungsto arsenate is higher than the exchange capacity of the material. Moreover, the uptake of Rb+ ion increases with increasing solution concentration and isotherm does not attain constancy even at 0.1 M concentration (Pig. 7). It is possible that ion-exchange as well as physical adsorption is responsible for a higher uptake of these ions. The membranes of these two gels are prepared using three different binding materials viz. parchment, polysty rene and aTaldite. It is found that the parchment based membranes do not swell and the aTaldite based (40*/.) also swell to a lesser extent as compared to polystyrene based IV (20*/.) membranes. This behaviour is attributed to the nature of binding material. The data on water content, porosity, electrolyte absorption and conductance for the two membranes having different binding materials follow the order: Parchment based . polystyrene based . araldite based membranes * membranes / membranes and the conductance values of the two membranes in various ionic forms exhibit the order; H+>Cs+>Rb"^K+'>Na+ / Li+ and Ba2+> Ca2+ (for titanium arsenate) H+> Rb+>Cs+>K+>Na+>Li+ and Ca2+>Ba2+ (for titanium tungstoarsenate) The conductance decreases with an increase in hydrated ionic radii of the ions, with the exception of Rb+, Ca2+ and Ba2+ in titanium tungstoarsenate membranes. It is further observed that conductance increases with ah increase in external solution concentration which is attributed to the fact that Donnan exclusion becomes less effective at higher solution concentrations thereby leading to a greater electrolyte absorption and a higher membrane conductance. Permeability of electrolytes through these membranes is measured by Willis method (Willis, G.M. 3nd Hartung, E.J., Trans-Paraday Soc. , 4.0., 520, 1944). It is found that membranes with araldite as binder have negligible permeability for the ions investigated whereas parchment based membranes show V maximum permeability. The diffusion of various salt solutions across the parchment and polystyrene based membrane* follow the order; K+>Na+>Li+ and Ba2+> Ca2+>Mg2+. The permeation of an electrolyte through a membrane depends upon the charge and size of the diffusing ion, charge on the mmbrane matrix, its porosity, exchange capacity of the membrane material and the nature of binder used (in the case of heterogeneous membranes). Permeability of these, ions decrease as the hydrated size of the ions increase. According to Eisenman-Sherry model, the permeability order observed in our case points towards a weak field strength of charge groupings in both the membranes. The thermodynamic parameters of the process of diffusion have also been calculated at 25°C by applying the theory of absolute reaction rates. It is observed that the values of AS^ are negative except that of lithium in the case of titanium tungstoarsenate membrane with parchment binder and also in the case of titanium arsenate membrane with polysty rene binder. The values of /s S' indicate the mechanism of flow; large positive /\ S' reflects the existence of a large zone of activation or breaking of bonds while a low negative value of entropy points to the fact that the activated state involved formation of some bond between the diffusing ion and vi membrane material. As such a positive value of A * obtained for Li+ in the case of polystyrene based titanium arsenate and parchment based titanium tungstoarsenate membranes indi cates that the diffusion of this ion across the membranes referred above, generates a greater region of disorder. The negative values of /\& obtained for other ions show that the permeating ions are partially immobilized in the memb rane structure or interfacial region. Membrane potential studies reveal that positive potentials are generated across parchment and polystyrene based membranes whereas negative potentials are observed across araldite based membranes of these two gels. The order of membrane potential parallels the order of permea bility of the membranes of these two ion exchangers contain ing different binding material. As far as electrolytes are concerned, no regular potential sequence is observed. The potential increases with time, rapidly attaining a maximum value after a certain interval and then falls off slowly in the case of parchment and polystyrene based membranes and remains constant for araldite based membranes. None of the membranes investigated show permselectivity and Nernstian response is not observed in any case. The transport number of permeating cations, t , in the membrane phase, permselectivity, Pss and apparent anionic mobility of these membranes has been calculated at various Vll concentrations and it is observed that the parameters increase with decreasing electrolyte concentration and reach a maximum value. The value of apparent anionic mobility decreases and reaches a minimum value at certain electrolyte concentration. Fixed charge density is determined by two different methods (Altug Hair and Kobatake). The value of fixed charge density is higher for parchment based in comparison to polystyrene based membranes. Among the membranes of these compounds with different binders, titanium tungstoarsenate membranes have higher value of charge density than titanium arsenate membranes. These studies also point towards the fact that membrane fixed charge density is not a constant quantity, but depends upon the concentration and nature of solutions in contact with the membranes. The titanium arsenate membrane with 40"/. araldite content is found to be selective for Pb + ions. The plot of potential vs log concentration of lead nitrate is perfectly -1 -«5 linear in the concentration range 10 to 10 M with a non- Nernstian slope of 33.33 mV per deCa.de of the concentration. The working concentration and pH range of the electrode assembly is 5xI0~ to 5xl0~ } M and 3.0 to 6.5 respectively. Potentials are reproducible, the response time is 50s, remains stable for 15 min. and the membrane can be used for four months It can be used in the continuous analysis of waste effluents Vlll containing lead ions in which the constituents may show major variations in concentration. For the assessment of selectivity of this membrane electrode for lead ions over other cations, the selectivity coefficient, k^°g, is determined for a large number of cations and anions by fixed interference method at 10 and 10~2 Mconcentration of interfering ions. It is observed that monovalent ions K+, NH*, Cs"1", Rb+ and Tl+ may cause interference if present in equivalent or larger concentra tions than primary ion . The electrode is quite tolerant of N0~ and CH,C00~ but Cl~ , SO?" and S2~ ions do interfere. It is further noticed that the membrane treated with cetyl pyridinium chloride (10 M) becomes immune to the effect of cationic surfactants in general and aLso exhibits better selectivity to lead ions. Smaller concentrations of anionic or non-ionic detergents do not cause any disturbance. It can be used in partially non-aqueous media (less than 30*/. ethanol) without observing any discrepancy in its properties. This electrode assembly haj3 also been used successfully as end point indicator in the titration of 25 ml of 10 M lead nitrate with 10" M sodium molybdate. 40*/. araldite based titanium tungstoarsenate membrane is found to be selective for rubidium ions. The potential vs log concentration of rubidium chloride plot is linear in concentration range 10 to 10"4 Mwith a non-Nernstian slope IX of 53.33 mV per decade of concentration. This electrode can be used to measure rubidium ion activity in the concentration range 10~ to 4x10"5 Mand the working pH range is 3-8. The response time of the measuring c&Ll is less than a min., potential remains stable for 10 to 15 min, and are reprodu cible. Membrane can be used for six months. The selectivity coefficient, kAPo|t , values as deter mined at a fixed concentration (10~5, lo"2 M) of interfering ions show that the bivalent and other polyvalent cations do not interfere even at higher concentration but monovalent cations may cause interference if their concentration approaches or exceeds the concentration of primary ions. The membrane on treataent with sodium dodecyl sulphate (10 M for 48 h), not only becomes immune to the effect of surfactant but also behaves as a better sensor for Rb ions. Electrode can also be used in partially non-aqueous media containing upto 25'/. non-aqueous content. The electrode has shown satisfactory results as end point indicator in the _p -2 titration of 5x10 N 12-tungstophosphoric acid with 10 '' N rubidium chloride solution.