**4. Removal of anions with Octolig® and metal derivatives of Octolig®**

Someinitialstudiesinourlaboratorywereconcernedwithremovalofnuisancespeciesbymeans of chromatography with Ferrilig, the iron(III) derivative of Octolig ® These species, existing as anions, were arsenate and arsenite, chromium as chromate, molybdenum as molybde‐ num(VI),andseleniumasseleniteandselenitespecies.Allfourarenuisancespeciesinthewestern United States as well as elsewhere. Molybdenum is an essential element, whose compounds are useful, but it is a nuisance in areas of molybdenum mining when mining runoff water or processingwaterinpondsbecomesadisposalproblem.Theinitialfocuswasonremovingarsenic species by chromatography, and a specific focus was on the iron(III) derivative of Octolig® (named "Ferrilig" ) because of the known insolubility of ferric arsenate [30-32].

The synthesis of Ferrilig (IV, Eqn 5), originally described [30] has been studied with view toward improving the amount of iron taken up. The synthesis is summarized (Eqn. 4-6) [31]. Octolig® (II) was treated with aqueous ferrous sulfate under a nitrogen atmosphere. The product (material III Eqn. 4, 5) after spontaneous oxidation and production of hydroxide ion (Eqn. 6) is termed Ferrilig (material IV). The structure of Octolig® was that given in the company literature [28].

$$\text{H}\_3\text{Si}\text{O}\text{--Si}-\text{CH}\_2\text{CH}\_2\text{CH}\_2\text{NHCH}\_2\text{CH}\_2\text{[NHCH}\_2\text{CH}\_2\text{]}\text{n NH}\_2\text{+Fe}^{2+} \rightarrow \text{III}\tag{4}$$

$$\text{OH} + \text{O}\_2 \rightarrow \text{0}\_3\text{Si-O-Si-CH}\_2\text{CH}\_2\text{CH}\_2\text{CH}\_2\text{NHCH}\_2\text{CH}\_2 \left[\text{NHCH}\_2\text{CH}\_2\text{\cdot}^\cdot\right] \text{Fe(III)}\tag{5}$$

$$\text{Fe}^+ + \text{O}\_2 + \text{H}\_2\text{O} \rightarrow \text{ 2OH}^-\tag{6}$$

Oxidation of the ferrous form (material III, green) to the ferric form (material IV, rust brown) occurs spontaneously in the presence of air, e.g., as the wet sample is standing exposed to the air. The role of coordination in reducing the oxidation potential of iron(II) is well known, and was noted by Moeller [33], a process that is enhanced when the coordinating agent is a chelating agent. Thus the oxidation potential for hydrated ferrous-ferric species is - 0.771V; whereas the value in the presence of oxalate ion is -0.02 V [34]. Octolig® has a plethora of chelating species, i. e., ethylenediimino moieties or extended ethylenediamines, that should be capable of lowering the oxidation potential of coordinated iron(II). Accordingly, the ease of oxidation should hardly be surprising. Nevertheless, it was surely interesting to note and watch. Species II was white, species III was green, and species IV was rust-brown.

were quantitatively removed by Ferrilig, Thorilig, or Octolig® [29]. Uranium is a contaminant of the mineral apatite in Florida that is a basis of the phosphate industry. Uranium can also contaminate sources of drinking water in certain areas of Colorado, and is a matter of concern

Gao and co-workers measured the absorption properties of an Octolig®- like material [13] by a batch and a flow methods. Quantitative reaction was reported, and the absorbing ability of the PEI-silica material followed the order of Cu2+ > Cd2+ > Zn2+ at a pH of 6-7 [13]. They also measured the saturated absorption uptake and reported values for copper(II) of 25.95 mg/g

**4. Removal of anions with Octolig® and metal derivatives of Octolig®**

(named "Ferrilig" ) because of the known insolubility of ferric arsenate [30-32].

Someinitialstudiesinourlaboratorywereconcernedwithremovalofnuisancespeciesbymeans of chromatography with Ferrilig, the iron(III) derivative of Octolig ® These species, existing as anions, were arsenate and arsenite, chromium as chromate, molybdenum as molybde‐ num(VI),andseleniumasseleniteandselenitespecies.Allfourarenuisancespeciesinthewestern United States as well as elsewhere. Molybdenum is an essential element, whose compounds are useful, but it is a nuisance in areas of molybdenum mining when mining runoff water or processingwaterinpondsbecomesadisposalproblem.Theinitialfocuswasonremovingarsenic species by chromatography, and a specific focus was on the iron(III) derivative of Octolig®

The synthesis of Ferrilig (IV, Eqn 5), originally described [30] has been studied with view toward improving the amount of iron taken up. The synthesis is summarized (Eqn. 4-6) [31]. Octolig® (II) was treated with aqueous ferrous sulfate under a nitrogen atmosphere. The product (material III Eqn. 4, 5) after spontaneous oxidation and production of hydroxide ion (Eqn. 6) is termed Ferrilig (material IV). The structure of Octolig® was that given in the

<sup>3</sup> 2222 22 22 2 -0 Si-O –Si – CH CH CH CH NHCH CH NH CH CH n NH + Fe III

( ) 2 3 2222 22 22 III + O -0 Si-O-Si -CH CH CH CH NHCH CH NH CH CH -


Oxidation of the ferrous form (material III, green) to the ferric form (material IV, rust brown) occurs spontaneously in the presence of air, e.g., as the wet sample is standing exposed to the air.

é ù

é ù e III ® ë û (5)

ë û ®

2 2 e + O + H O 2 OH ® (6)

2+

F

(4)

and 50.01 mg/g, respectively, for static and dynamic conditions [13].

for water supplies for small towns [29].

132 Column Chromatography

company literature [28].

(

IV)

(II)

Considering the effectiveness of Ferrilig, the study of other metal derivatives ("metalloligs") was effected using what may be described as facile syntheses. The metals used were cop‐ per(II), cobalt(II), nickel(II), manganese (II), and thorium(IV) [32]. An exhaustive study can not be claimed, e.g., for all metalloligs and all anions. But all six metalloligs exhibited 99% remov‐ al of arsenic by means of column chromatography using 280 x a10-3 ppm As as Na2HAs04 [32].Other anionswere testedusingvariousmetalloligs, andquantitative removal(98-99%)was achieved for nitrate, nitrite, phosphate, sulfate, and fluoride ions in deionized water [31, 32, 35].

A *standard test* for removal ions by chromatography involved the following: A Spectra/chron peristaltic pump was used to deliver aqueous samples to a chromatography column, 2 cm (id) by 31cm and equipped with a glass frit and a Teflon stopcock. The column was packed with about 22 cm of Octolig® or other solid. Before packing, the solid was suspended in water, swirled, and the fines were decanted, a process that was repeated until no fines were observed. Water samples were chromatographed using a rate of 10 mL/min. Usually, the first three or four 50-mL aliquots of effluent were discarded, and later ones were used for analysis (Table 4). Total dissolved solids were measured, and used as a guide to assess a state of equilibrium [29-31, 35-38].



In a subsequent study, some attention was focused on the use of Cuprilig, obtained by a truly facile synthesis by shaking a suspension of Octolig® in deionized water and a standard solution of copper sulfate in deionized water. Cuprilig was tested for removal of perchlorate ion, which is a serious problem in certain areas, most notably in Rialto, California where one source of well water contained 10,000 ppb perchlorate. This remarkable concentration was probably a consequence of a plume of contaminated water, owing to proximity to a facility that produced ammonium perchlorate, the propellant for the sidewinder missile [37]. Obvi‐ ously it was of interest to determine whether a metallolig, e.g., Cuprilig, could remove perchlorate ion from aqueous solutions using column chromatography, and it was a success [37]. Using perchlorate solution, the effluent was below detection limit, i.e., <1 ppb. But success also raised the question of a "control" experiment, one kind being to eliminate the role of the metal ion, and use only Octolig®. Doing so, the results were equally good, i.e., > 99% removal with Octolig® alone using deionized water or well water [37].

Clearly a different model for understanding the removal process was in order. And the current hypothesis is that nitrogens are protonated at low pH values probably as well, even somewhat above the pH at which Octolig® should not used. A working hypothesis consistent with Figure 5 is that at least two important considerations are involved: (1) The nitrogens must be proto‐ nated, and (2) species to be removed must be anions or if weak acids convertible to anions [39].

Further studies [35, 37, 38] have found that simple anions can be quantitatively removed by Octolig ® at reasonable values of pH, and the order of removal is consistent with an ionic model of anions being attracted to protonated moieties of the Octolig® as represented (Figure 5).

**5. Control of pharmaceuticals and other large anions**

**Figure 6.** Maximum per cent removed as a function of the pKa for the compound under study [47]

will be considered in succession.

the environment in the form of urine and feces [41].

**5.1. Pharmaceuticals**

Three diverse groups of compounds can be usefully considered for their impact, but the need to be able to separate them from aqueous solution: Pharmaceuticals, BPA, and food dyes. These

Chromatographic Separations with Selected Supported Chelating Agents

http://dx.doi.org/10.5772/55521

135

The range of anions that can be quantitatively removed by Octolig® and column chromatog‐ raphy include some significant pharmaceuticals. These substances are chemical compounds (inorganic or organic) that can be used in the diagnosis, mitigation, treatment, or prevention of a disease [41,42]. Unfortunately, because of their usefulness and the magnitude of produc‐ tion they can represent a disposal problem [41-43]. Pharmaceuticals have an impressively wide range of applications – human medicine, veterinary medicine, aquaculture, livestock produc‐ tion, and agriculture. [43]. A recent review [42] quoted a statistic that of the 16,200 tons of antibiotics produced in the United States in 2000, about 70-% was used for livestock [40]. Unfortunately, about 75% of the antibiotics involved was not absorbed and was delivered to

On the positive side, a study using Octolig® as a separating agent showed that Amoxicillin (Figure7),averypopularantibioticintheUnitedStateswasamongthosequantitativelyremoved [44], as noted in Table 5. This a result that is consistent with the information provided in Figure 6. It is also noteworthy that results for DI (deionized ) water and well water from the Floridan Aquifer were similar. The lack of a matrix effect, at least for these two solvents may be promis‐

**Figure 5.** Proposed structure of Octolig® -anion (A-) interaction [38]. Reproduced with permissionof the publisher

A range of compounds can be quantitatively removed provided the pKa of the material is less than about 8 [40]. Above that value, the per cent removal was less than 20% [40]. This is demonstrated in Figure 6. It may be presumed that for those compounds having pKa values greater than 7+, the anion concentration vs. the degree of protonation of Octolig® reaches an unfavorable balance. One substance of special interest is BPA (*vide infra*).

Compounds in order of increasing pKa values (in parentheses): Amoxicillin (2.4), Eosin Y (2.7), Lissamine Green (~3), Erythrosine (3.6), Rose Bengal (3.9), 4-nitrophenol (7.15), 2-nitrophenol (7.22), 3-nitrophenol (8.36), 4-tert-butylphenol (10.16) 4-isoproplphenol (10.19). From [47] used with permission

**Figure 6.** Maximum per cent removed as a function of the pKa for the compound under study [47]
