Removal of Zinc From Phosphate Solutions by Anion-Exchange

In plant nutritional studies. it is often necessary to prepare nutrient solutions that are free of a particular micronutrient. It is occasionally possible to obtain batches of nuitrient salts from commercial sources that are sufficiently loyv in a given micronutrient to be used in such1 deficiency studies. But as pointed out by Hewitt (2), it is muich more advisable in these instances, to use appropriate purification methlods in order to obtain consistently good results. Altlhough the amount and type of contamination are higlhly variable, phosphates quite generally contain significant amounts of the micronutrients, especially zinc (2). Removing zinc from phosphate solutions is particularly difficult because of the existence of stable ion pairs. Commonly used methods (coprecipitation and organic extraction following chelation) rely primarily on the presence of ionic forms of zinc. In 1953, Kraus and Moore (4) described the anion-exchange behavior of the divalent transition elements Mn to Zn in clhloride media. Their resuilts slhowed that all of the elements studied, except Ni, formed negatively charged chloride complexes which could be adsorbed on an anion-exchange resin. The zinc-chloride complex was strongly adsorbed from 2 M HCI. This adsorption is the basis for the separation procedure described here. The anion-exchange resin used was AGI-X8 in the chloride form, 200 to 400-mesh (Dowex 1 reprocessed by Bio-Rad Laboratories, Berkeley. California). The resin was packed into a 12 X 600-mm polyethylene column witlh a 100-ml polyethvlene reservoir attaclhed to the upper end. Both ends of the resin column were plugged with Dacron polyester wool. Tlle columns were pretreated with 100 ml concentrated reagelnt-grade HCI followed by 400 ml glass-distilled water. Concentrated reagentgrade H3PO was diluted with glass-distilled water and concentrated HCI, giving a solution that was 6 M H3PO4 and 2 M HCl. Before this solution was placed on the column, approximately 10 ml 2 M HG1 was allowed to pass into the column. The H3PO4HCI solution was passed through the column at 0.1 to 0.2 mil/min, collected, neutralized with purified KOH, and diluted to a convenient concentration. Preliminary experiments with 03Zn suggested that the procedure was very effective in removing Zn. An estimate was desirable, however, of the amount of Zn expectable in a nutrient solution containing plhosplhate purified in this manner. A microbiological assay for Zn was therefore used. The test orgnni-m was Rhizobiutmz meliloti, strain SU 47 (CSIRO. Division of Plant Industry, Canberra, Atustralia). The basic culture soltution had the following composition: mannitol, 15 mM; potassium phosphatle (pH 7), 12 mM; potassitum gltutamate (pHf 7), 4 mM; NHT1Cl, 4 mM; MgSO4, 1 nmM; CiCl.,. 02 mM: FrCl, 10 pm; MnCl., 0.1 gm; H:iBO,, 0.1 mM :CuSO,. 0.01 gM: NaXMoO4. 0.01 gM; Co(NO:, 0.005 Mm; biotin, 025 ppm; calcium pantothenate, 0.10 ppnm; and thiamine hydrochloride. 0.10 ppm. The final solution was adiusted to pH 7.0 with purified KOH or redistilled HCI. The nutrients, otlher than phosphate, were purified as follows: mannitol by passage througlh a 25 X 250-mm column of cation-exclhange resin (AG50W-X8, Bio-Rad Laboratories. Berkelev, California); potassium glutamate, MgSO4, CaClI, and KOH by Mg(OH)2 coprecipitation (5); NH,CI from redistilled NH3 and HCl; FeCl3 by etlher extraction (1); H3BO,. CuSO4, and Na..MoO4 by recrystallization; and MnCl2 and Co(NO3 ) from the spectrograplhically standardized metals (Jolhnson Mattlhev, Ltd., London) dissolved in redistilled Ha or HNO3 and diluting with glass-distilled water. The vitamins wvere used withlout further purification. The Zn variables were added as ZnSO4, purified by recrystallization. Polycarbonate flasks (50 ml) containing 10 ml of the autoclaved culture solution were inoculated with 1 drop of a suspension of Zn-deficient rhizobia. The cultures were incubated for 72 hr at 310 with continuous shaking. Growth was measured as absorbance at 600 mMA in cuvettes lhaving a 1-cm liglht patlh. The average value of 3 replications for each Zn level is reported. Table I slhows the effect of added Zn on the growvtl of the rlhizobia. Extrapolation of the 3 lowest levels of added Zn back to zero absorbance gives a Zn contamination level of approximately 10-10 M in a solution containing 12 mm plhosphate. Considering that all of the other nutrients in the culture solution would undoubtedly supply some Zn and that the phosphate level is unduly high for plant culture, it is likely that phosphate, purified as described, would supply no more than 10-11 m Zn in