Synthesis of catechol-based adsorbents using microwave radiation and selective separation technology for metal and metalloids

Abstract

Embodiments relate to an adsorbent comprising an aminated solid support functionalized with at least one catechol-based compound, such as catechol, nitro-catechol, and / or pyrogallol. The adsorbent may be configured to be used for germanium recovery from dilute acidic sources. The adsorbents may be synthesized using microwave radiation. Embodiments further relate to methods of using catechol-based adsorbents, particularly to using adsorbents in fixed-bed columns for germanium adsorption.

Description

Atty. Ref. No. 0073605-000829 SYNTHESIS OF CATECHOL-BASED ADSORBENTS USING MICROWAVE RADIATION AND SELECTIVE SEPARATION TECHNOLOGY FOR METAL AND METALLOIDS CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] The present application claims priority to U.S. Provisional Patent Application No. 63 / 509,922, filed on June 23, 2023, the entire contents of which is incorporated by reference. FIELD OF THE INVENTION

[0002] Embodiments relate to catechol-based adsorbents and methods of making thereof, particularly to adsorbents synthesized using microwave radiation and configured to adsorb germanium. Embodiments further relate to methods of using catechol-based adsorbents, particularly to using adsorbents in fixed-bed columns for germanium adsorption. BACKGROUND OF THE INVENTION

[0003] Germanium is a valuable and rare metalloid, and is considered to be a critical element due to its application in infrared systems, optical fibers, semiconductors, solar panels, and polymer catalysis. As the demand for germanium continues to grow, driven by emerging applications such as infrared optics for car safety, the expansion of 5G networks, and the use of multi-junctional solar cells on terrestrial surfaces, the need for efficient and profitable production processes for germanium recovery is becoming increasingly urgent. Currently, germanium is primarily recovered as a byproduct from zinc refineries and coal fly ash, yet worldwide only a small fraction of germanium (3-5%) is recovered from these sources. Moreover, the low content of germanium and high concentration of impurities in these sources create technological and economic challenges for germanium recovery. Additionally, recycling of electronic waste resultsAtty. Ref. No. 0073605-000829 in dilute germanium solutions after leaching. Consequently, there is a pressing need for the development of effective and economically feasible production processes that can recover germanium from low-content sources. SUMMARY

[0004] A production process based on solid-phase extraction (SPE), primarily using functionalized adsorbents that selectively adsorb germanium, can offer efficient extraction at low concentration levels. Various adsorbents have been utilized for germanium adsorption, but only a few of these adsorbents have demonstrated high capacities. Further, they lack selectivity for germanium over competing ions. Cellulose grafted with di(2-hydroxyethyl)amine has previously exhibited a germanium adsorption capacity of 130.7 mg / g at pH 8, with germanium selectivity factors ranging from 5 against Te to 12 against As, Sb, Si, and Ni. On the other hand, chitosan grafted with poly (hydroxyethyl methacrylate) has demonstrated a maximum capacity of 73.91 mg / g at pH 4, but it was not selective against Pb. Although N-methylglucamine, another hydroxyl-based ligand, has exhibited a high capacity for germanium adsorption, it lacked selectivity. Nevertheless, several commercial adsorbents containing the N-methylglucamine ligand, such as Purolite S108, Amberlite IRA743, and Diaion CRB 02, are available.

[0005] Catechol, a dihydroxyl ligand, has previously exhibited remarkable selectivity for Ge in solid-phase extraction. Catechol-functionalized nano-silica has demonstrated a maximum capacity of 6.08 mg / g at pH 4.5, and it exhibited selectivity for germanium over B and Te with selectivity factors of 2.1 and 74, respectively. Similarly, catechol-formaldehyde resin has shown selectivity for germanium over Si and Zn with selectivity factors of 28 and 5, respectively (at pH 6 and 1g resin / L), but it had a capacity of less than 2 mg / g.Atty. Ref. No. 0073605-000829

[0006] Further, existing adsorbents have been tested in batch systems, but industrial application involves dynamic column adsorption. Fixed-bed adsorption columns find extensive application in removing target elements from aqueous solutions. By using ligand-functionalized adsorbents selective to target ions, it is possible to extract and recover the target element from a complex solution. The performance of the fixed-bed adsorber depends on the composition (i.e., ions present, concentrations, and pH) of the inlet solution and the operating parameters (i.e., bed height, bed volume, and flow rate). By optimizing these conditions, selective separation of target ions (e.g., germanium) can be achieved.

[0007] Embodiments relate to an adsorbent comprising an aminated solid support functionalized with at least one catechol-based compound, such as catechol, nitro-catechol, and / or pyrogallol. The adsorbents may be synthesized using a method involving microwave radiation. Microwave heating operates by aligning the dipoles of the material in an external field by the excitation produced by microwave electromagnetic radiation. Microwave heating offers the advantages of high heating rates, homogenous heating with high-energy efficiency, and shorter synthesis time. Embodiments further relate to using said adsorbents in a fixed-bed system to remove target elements, such as germanium, from aqueous solutions.

[0008] In exemplary embodiments, an adsorbent comprises an aminated solid support; and at least one catechol-based compound, wherein the at least one catechol-based compound is selected from the group consisting of catechol, nitro-catechol, pyrogallol, and mixtures thereof.

[0009] In some embodiments, the adsorbent is synthesized by microwaving the aminated solid support and the at least one catechol-based compound.

[0010] In some embodiments, the aminated solid support comprises a base material, wherein the base material is polystyrenic beads or silica.Atty. Ref. No. 0073605-000829

[0011] In some embodiments, the aminated solid support comprises surface amines and wherein the ratio of surface amines to catechol-based compound is between 4:1 and 6:1.

[0012] In some embodiments, the adsorbent is selective for germanium in the presence of competing ions.

[0013] In some embodiments, the adsorbent has a selectivity factor for germanium greater than 3.6.

[0014] In exemplary embodiments, a method of synthesizing an adsorbent comprises selecting an amount of an aminated solid support and an amount of at least one catechol-based compound; mixing the selected amounts of aminated solid support and at least one catechol-based compound; and microwaving the mixture to form a microwaved product.

[0015] In some embodiments, the method further comprises filtering the microwaved product; and / or washing the microwaved product; and / or drying the microwaved product.

[0016] In some embodiments, the method further comprises reacting the microwaved product with sodium borohydride.

[0017] In some embodiments, the aminated solid support is provided in an aqueous medium prior to mixing the selected amounts of aminated solid support and at least one catechol-based compound.

[0018] In exemplary embodiments, a method of using an adsorbent to capture germanium from a source, the adsorbent comprising an aminated solid support; and at least one catechol-based compound, wherein the at least one catechol-based compound is selected from the group consisting of catechol, nitro-catechol, pyrogallol, and mixtures thereof, the method comprising: exposing the adsorbent to a source comprising germanium; and adsorbing germanium from the source.Atty. Ref. No. 0073605-000829

[0019] In some embodiments, the method further comprises releasing the adsorbed germanium with at least one mineral acid.

[0020] In some embodiments, the aminated solid support comprises a base material, wherein the base material polystyrenic beads or silica.

[0021] In some embodiments, the adsorbent is selective for germanium in the presence of competing ions.

[0022] In some embodiments, the competing ions are selected from the group consisting of Be(II), Mg(II), Al(III), Sc(III), Mn(II), Fe(III), Cu(II), Zn(II), Ga(III), As(IV), Cd(II), La(III), Yb(III), Pb(II), Th(IV), UO22+, and mixtures thereof.

[0023] In some embodiments, the adsorbent has a selectivity factor for germanium greater than 3.6.

[0024] In some embodiments, the source is selected from the group consisting of acid mine drainage, mine tailings leachate, zinc refineries residue, and leached solutions.

[0025] In some embodiments, the leached solutions are selected from the group consisting of coal fly ash, fiber optics, and electronics.

[0026] In exemplary embodiments, a method of using an adsorbent to capture germanium from a source, the adsorbent comprising an aminated solid support; and at least one catechol-based compound, wherein the at least one catechol-based compound is selected from the group consisting of catechol, nitro-catechol, pyrogallol, and mixtures thereof, the method comprising providing at least one fixed bed comprising the adsorbent; passing the source through the at least one fixed bed; and adsorbing germanium from the source.

[0027] In some embodiments, the method comprises providing a series of fixed beds comprising the adsorbent.Atty. Ref. No. 0073605-000829

[0028] In some embodiments, the at least one fixed bed is provided in an adsorption column.

[0029] In some embodiments, the source is passed through the at least one fixed bed at a flow rate less than 1 mL / min.

[0030] In some embodiments, the method further comprises passing a mineral acid through the at least one fixed bed to release the adsorbed germanium.

[0031] In some embodiments, the adsorbent retains greater than 70% capacity after five adsorption-desorption cycles. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The above and other objects, aspects, features, advantages and possible applications of embodiments of the present innovation will be more apparent from the following more particular description thereof, presented in conjunction with the following drawings. Like reference numbers used in the drawings may identify like components.

[0033] FIG. 1 shows an exemplary method of making an exemplary adsorbent.

[0034] FIG. 2 shows an exemplary method of using an exemplary adsorbent to capture germanium (Ge) from a source.

[0035] FIG. 3 is a schematic illustration of microwave-based functionalization of aminated solid support.

[0036] FIG. 4 shows a Fourier-transform infrared spectroscopy (FTIR) curve for a polystyrenic bead (e.g., A110), A=Cat, and A-Cat.

[0037] FIG. 5 shows a FTIR curve for a polystyrenic bead (e.g., A110), A=Cat-N, and A-Cat-N.

[0038] FIG. 6 shows a FTIR curve for a polystyrenic bead (e.g., A110), A=Py, and A-Py.

[0039] FIG. 7 shows a FTIR curve for Silica-NH2, S=Cat, and S-Cat.

[0040] FIG. 8 shows a FTIR curve for Silica-NH2, S=Cat-N, and S-Cat-N.Atty. Ref. No. 0073605-000829

[0041] FIG. 9 shows scanning electron microscope (SEM) images of A110, A-Cat, and A-Cat-N at 200,000x.

[0042] FIG. 10 shows energy-dispersive x-ray spectroscopy (EDS) elemental mapping of A-Cat after adsorption of Ge.

[0043] FIG. 11 shows EDS elemental mapping of A-Cat-N after adsorption of Ge.

[0044] FIG. 12 is a graph showing Ge adsorption onto A-Cat, A-Cat-N, and A-Py at different initial pH values (initial Ge concentration: 55 mg / L).

[0045] FIG. 13 is a graph showing initial and equilibrium pH in the of A-Cat, A-Cat-N, and A- Py as a function of different initial pH values (initial Ge concentration: 55 mg / L).

[0046] FIG. 14 is a graph showing Ge-ligand complexation in a solution as a function of pH (Ge: 0.01 M, ligand: 0.03 M).

[0047] FIG. 15 is a graph showing distribution coefficients (Kd) of various elements in competitive-ion adsorption with A-Cat, A-Cat-N, and A-Py, initial conditions: 20 mg / L each element, pH 1, pH 2, and pH 3 (data represents the average of triplicates).

[0048] FIG. 16 shows selectivity factors (SF) of A-Cat and A-Cat-N for Ge against other elements.

[0049] FIG. 17 is a graph showing linear correlations (with intercept forced to zero) between the log Kadsfor the functionalized adsorbents and logK(MOH) of the metal cations at pH 2 (S represents the slope of the linear fit and the R2is the regression coefficient).

[0050] FIG. 18 is a graph showing linear correlations (with intercept forced to zero) between the log Kadsfor the functionalized adsorbents and logK(MOH) of the metal cations at pH 3 (S represents the slope of the linear fit and the R2is the regression coefficient).Atty. Ref. No. 0073605-000829

[0051] FIG. 19 is a graph showing Ge adsorption as a function of increasing equilibrium Ge concentration at pH 3 for A-Cat, A-Cat-N, and A-Py (lines show the Langmuir isotherm model fit).

[0052] FIG. 20 is a graph showing the effect of contact time on the adsorption of Ge with adsorbent A–Cat at pH 3 and 80 mg / L Ge (lines represent the PFO and PSO models).

[0053] FIG. 21 is a graph showing the effect of contact time on the adsorption of Ge with adsorbent A–Cat-N at pH 3 and 80 mg / L Ge (lines represent the PFO and PSO models).

[0054] FIG. 22 is a graph showing the effect of contact time on the adsorption of Ge with adsorbent A–Py at pH 3 and 80 mg / L Ge (lines represent the PFO and PSO models).

[0055] FIG. 23 is a graph showing the Ge adsorption and desorption cycle for the different adsorbent synthesized ((A): Adsorption step, (D): Desorption step).

[0056] FIG. 24 is a graph showing an x-ray photoelectron spectroscopy (XPS) survey spectrum for a A-Cat adsorbent before and after Ge(IV) adsorption.

[0057] FIG. 25 is a graph showing an XPS survey spectrum for a A-Cat-N adsorbent before and after Ge(IV) adsorption.

[0058] FIG. 26 is a graph showing an XPS survey spectrum for a A-Py adsorbent before and after Ge(IV) adsorption.

[0059] FIG. 27 is a graph showing Ge 3d XPS spectra of the adsorbents after Ge adsorption.

[0060] FIG. 28 is a graph showing breakthrough curves (adsorption) for a Ge(IV) solution in columns packed with A-Cat, A-Cat-N, and A-Py. (Co= 10 mg / L Ge(IV); I =0.01 M KCl; pH = 3; Q = 30 BVH).

[0061] FIG. 29 is a graph showing breakthrough curves (desorption) for a Ge(IV) solution in columns packed with A-Cat, A-Cat-N, and A-Py.Atty. Ref. No. 0073605-000829

[0062] FIG. 30 is a graph showing adsorption and elution curves for multi-element solution for A-Cat (adsorption at pH 2).

[0063] FIG. 31 is a graph showing adsorption and elution curves for multi-element solution for A-Cat (desorption after pH 2 adsorption).

[0064] FIG. 32 is a graph showing adsorption and elution curves for multi-element solution for A-Cat (adsorption at pH 3).

[0065] FIG. 33 is a graph showing adsorption and elution curves for multi-element solution for A-Cat (desorption after pH 3 adsorption).

[0066] FIG. 34 is a graph showing adsorption and elution curves for multi-element solution for A-Cat-N (adsorption at pH 2).

[0067] FIG. 35 is a graph showing adsorption and elution curves for multi-element solution for A-Cat-N (desorption after pH 2 adsorption).

[0068] FIG. 36 is a graph showing adsorption and elution curves for multi-element solution for A-Cat-N (adsorption at pH 3).

[0069] FIG. 37 is a graph showing adsorption and elution curves for multi-element solution for A-Cat-N (desorption after pH 3 adsorption).

[0070] FIG. 38 is a graph showing adsorption and elution curves for multi-element solution for A-Py (adsorption at pH 2).

[0071] FIG. 39 is a graph showing adsorption and elution curves for multi-element solution for A-Cat-N (desorption after pH 2 adsorption).

[0072] FIG. 40 is a graph showing adsorption and elution curves for multi-element solution for A-Py (adsorption at pH 3).Atty. Ref. No. 0073605-000829

[0073] FIG. 41 is a graph showing adsorption and elution curves for multi-element solution for A-Py (desorption after pH 3 adsorption).

[0074] FIG. 42 shows a table comparing the three adsorbents (A-Cat, A-Cat-N, and A-Py) in multi-element fixed-bed adsorption / desorption in terms of Ge(IV) selectivity.

[0075] FIG. 43 is a graph showing breakthrough curves for Ge(IV) adsorption onto A-Cat at different flow rates (Co = 30 mg / L Ge(IV)) (pH 3, I = 0.01 M KCl). The dashed line represents the MDR model.

[0076] FIG. 44 is a graph showing breakthrough curves for Ge(IV) adsorption onto A-Cat at different inlet Ge(IV) concentration (Q = 30 BVH (= 0.5 mL / min)). The dashed line represents the MDR model.

[0077] FIG. 45 is a graph showing fixed-bed column adsorption-desorption cycles of A-Cat for an adsorption step (Co = 30 mg / L Ge(IV), pH 3, I = 0.01 M KCl, Q = 30 BVH).

[0078] FIG. 46 is a graph showing fixed-bed column adsorption-desorption cycles of A-Cat for a desorption step.

[0079] FIG. 47 is a graph showing synthetic Zn refinery residue (509 mg / L Fe(III), 1000 mg / L Cu(II) and Zn(II) each, 50 mg / L Co(II), Ni(II), and Ge(IV) each, and 40 mg / L Ga, pH 2) adsorption with A-Cat.

[0080] FIG. 48 is a graph showing synthetic Zn refinery residue (509 mg / L Fe(III), 1000 mg / L Cu(II) and Zn(II) each, 50 mg / L Co(II), Ni(II), and Ge(IV) each, and 40 mg / L Ga, pH 2) desorption step with A-Cat.

[0081] FIG. 49 is a graph showing synthetic Zn refinery residue (509 mg / L Fe(III), 1000 mg / L Cu(II) and Zn(II) each, 50 mg / L Co(II), Ni(II), and Ge(IV) each, and 40 mg / L Ga, pH 2) adsorption step with S108.Atty. Ref. No. 0073605-000829

[0082] FIG. 50 is a graph showing synthetic Zn refinery residue (509 mg / L Fe(III), 1000 mg / L Cu(II) and Zn(II) each, 50 mg / L Co(II), Ni(II), and Ge(IV) each, and 40 mg / L Ga, pH 2) desorption step with S108.

[0083] FIG. 51 is a graph showing the Ge-ligand complexation modeling using log KMLestimated using PHREEQC-PEST for nitro-catechol. The markers are Ge-ligand complexation % converted from pH-adsorption data.

[0084] FIG. 52 is a graph showing the Ge-ligand complexation modeling using log KMLestimated using PHREEQC-PEST for pyrogallol. The markers are Ge-ligand complexation % converted from pH-adsorption data.

[0085] FIG. 53 is a graph showing the Ge-ligand complexation modeling using log KMLestimated using PHREEQC-PEST for catechol. GeL and GeL2are pseudo-species. The markers are Ge-ligand complexation % converted from pH-adsorption data.

[0086] FIG. 54 is a graph showing the correlation between ^^^^�and log K(MOH).

[0087] FIG. 55 is a graph showing the estimation of ^^^^based on log KMLavailable and ^^^^�values estimated from log K(MOH).

[0088] FIG. 56 is a graph showing the comparison log KMLestimated using LFER and log KMLfor catechol and nitro-catechol with different ^^^^value.

[0089] FIG. 57 is a table providing the log KML(25°C) for three ligands and different elements and estimated using LFER (for simplicity, the charges are not shown).

[0090] FIG. 58 is a graph showing adsorption complexation data from adsorption (markers) and aqueous complexation modeling (dashed-lines) using PHREEQC for a multi-element mix solution (20 mg / L each) using log KML(see FIG. 57) for catechol; (B) nitro-catechol; (C) pyrogallol.Atty. Ref. No. 0073605-000829

[0091] FIG. 59 is a graph showing adsorption complexation data from adsorption (markers) and aqueous complexation modeling (dashed-lines) using PHREEQC for a multi-element mix solution (20 mg / L each) using log KML(see FIG. 57) for nitro-catechol.

[0092] FIG. 60 is a graph showing adsorption complexation data from adsorption (markers) and aqueous complexation modeling (dashed-lines) using PHREEQC for a multi-element mix solution (20 mg / L each) using log KML(see FIG. 57) for pyrogallol.

[0093] FIG. 61 is a graph showing adsorption complexation data from adsorption (markers) and aqueous complexation modeling (dashed-lines) using PHREEQC for nitro-catechol using log KMLfrom FIG. 57 but replacing log KMLfor Fe(III) from reported value (log KML= 17.1 ) to LFER estimated value (log KML= 16.32). DETAILED DESCRIPTION

[0094] The following description is of exemplary embodiments and methods of use that are presently contemplated for carrying out the present invention. This description is not to be taken in a limiting sense but is made merely for the purpose of describing the general principles and features of various aspects of the present invention. The scope of the present invention is not limited by this description.

[0095] Embodiments relate to an adsorbent comprising an aminated solid support functionalized with at least one catechol-based compound. It is understood that an “aminated solid support” is a base material having amine on its surface. For example, the aminated solid support can include any solid support that has amine on the surface. It is further understood that a “catechol-based compound” is a compound with a base formula C6H6O2with optional, additional functional groups. As seen in FIG. 3, in exemplary embodiments, the at least one catechol-based compound may be selected from the group consisting of catechol (A-Cat), nitro-catechol (A-Cat-N),Atty. Ref. No. 0073605-000829 pyrogallol (A-Py), and mixtures thereof. In other embodiments, the catechol-based compound can include A-Cat, A-Cat-N, or A-Py.

[0096] It is contemplated that the base material of the aminated solid support can be any inert material that supports amine on its surface. In exemplary embodiments, the base material may be selected from the group consisting of polystyrenic beads and silica. The base material may be of any shape and / or structure, including but not limited to spherical, powder, flakes, nano-particles, or membranes. In preferred embodiments, the base material is porous. The porosity of the base material can be microporous (e.g., pores less than 2 nm) or mesoporous (e.g., pores between 2 and 50 nm).

[0097] It is contemplated that an adsorbent can have one or more supports. For instance, an adsorbent can include plural supports comprising a first support and a second support. The first support can be the same as or different from the second support. For example, an adsorbent can have a first support that is polystyrene and a second support that is silica.

[0098] As seen in FIG. 1, a method of synthesizing the adsorbent comprises selecting an amount of an aminated support and an amount of at least one catechol-based compound, and mixing the aminated solid support with at least one catechol-based compound. In exemplary embodiments, the aminated solid support may be provided in an aqueous medium, and the aqueous medium may be mixed with the at least one catechol-based compound. For example, the aqueous medium may be an organic solution, such as a 10% acetic acid in methanol solution, but other mediums may be used. FIG. 3 demonstrates an exemplary adsorbent structure with polystyrene as solid support (A-Cat, A-Cat-N, and A-Py). It is contemplated that the catechol-based compound may be present in the resulting mixture in a surface amine to compound ratio ranging from 4-6 to 1, preferably 4.5-5.5 to 1, most preferably 5-5.1 to 1.Atty. Ref. No. 0073605-000829

[0099] The method further comprises microwaving the mixture. It is contemplated that microwaving the mixture may replace traditional methods, such as heating via reflux. Microwaving the mixture may reduce the required time for heating from, e.g., 24 hours, to mere minutes, while also achieving more reaction yield. In some embodiments, microwaving the mixture may include power using 20% of a 1140 W rated microwave, and the duration for microwave heating may be 4 minutes. It is contemplated that microwave heating operates by aligning the dipoles of the components in an external field by the excitation produced by microwave electromagnetic radiation. It is understood that heating offers the advantages of high heating rates, homogenous heating with high-energy efficiency, and shorter synthesis time.

[0100] As shown in FIG. 1, after the mixture is microwaved, the microwaved product may then be subjected to at least one or more additional steps. In one embodiment, the microwaved product may be filtered to separate the solid support from an aqueous mixture. For example, the microwaved product may be removed from aqueous mediums. In another embodiment, the microwaved product may be washed (e.g., with an organic solvent, such as methanol, ethanol, etc.) to remove anu unreacted reactants. In another embodiment, the microwaved product may be dried. It is contemplated that drying the microwaved product may be accomplished using any suitable means, such as air drying, heating, etc.

[0101] In exemplary embodiments, the microwaved product may further be reacted with sodium borohydride (NaBH4). It is understood that reacting the product with NaBH4reduces carbon–nitrogen double bonds (C=N) and solves the potential issue of hydrolysis of C=N in an aqueous medium.

[0102] It is contemplated that a grafting yield may be dependent on the starting aminated solid support. The term “grafting yield” means the percent of surface amine bonded to aAtty. Ref. No. 0073605-000829 catechol-based compound. In other words, the conversion of amines on the surface (surface- amine) to ligands (surface-amine-ligand). For example, a grafting yield of 44-70% may be achieved for the catechol-based components with polystyrenic bead, and a grafting yield of more than 26.6% may be achieved for the catechol-based components with silica.

[0103] In exemplary embodiments, the adsorbent may be used to adsorb germanium (e.g., Ge(IV)). In a preferred embodiment, the adsorbent is selective for germanium. The term “selective for germanium” or “selectivity for germanium” is the preference for the adsorbent to adsorb germanium as opposed to other competing ions. It is contemplated that the adsorbent is selective for germanium in the presence of equal or different amounts of competing ions, including but not limited to, Be(II), Mg(II), Al(III), Sc(III), Mn(II), Fe(III), Cu(II), Zn(II), Ga(III), As(IV), Cd(II), La(III), Yb(III), Pb(II), Th(IV), and / or UO22+.

[0104] It is contemplated that the adsorbent’s selectivity factor for germanium is at least 3.6. It is further contemplated that the adsorbent may demonstrate complete selectivity (e.g., the adsorbent may not adsorb competing ions at all). The term “selectivity factor” is the ratio of a target ion to remaining / competing ions.

[0105] In exemplary embodiments, as shown in FIG. 2, the adsorbent may be used to adsorb germanium from a germanium-containing aqueous source / solution such as acid mine drainage, mine tailings leachate, zinc refineries residue, and leached solutions of coal fly ash, fiber optics, and electronics. The method can therefore involve introducing or exposing an embodiment of the adsorbent to the solution, followed by allowing the metal or metal ion to adsorb to the adsorbent via metal complexation. The solution may be acidic. It is contemplated that the pH of the solution may influence the adsorption capacity of the adsorbent. Thus, the method may also involve adjusting pH of the solution to influence adsorption capacity of theAtty. Ref. No. 0073605-000829 adsorbent. It is further contemplated that that the solution may be dilute (for example, less than 1g / L Ge).

[0106] The adsorbents may have an adsorption capacity between 25-40 mg / g. For example, in embodiments wherein the catechol-based component is A-Cat, the adsorption capacity may be 29.8 mg / g. In embodiments wherein the catechol-based component is A-Cat-N, the adsorption capacity may be 39.1 mg / g. In embodiments wherein the catechol-based component is A-Py, the adsorption capacity may be 37.1 mg / g. It is contemplated that the adsorption capacity are dependent on the grafting yields.

[0107] In some embodiments, the adsorbent may be used in a fixed bed adsorption process. In a fixed bed adsorption process, the aqueous solution to be treated passes through at least one bed packed with an adsorbent. The fixed bed adsorption process therefore includes two phases – a moving aqueous solution and a stationary solid phase of the packed adsorbent. As the solution passes through the at least one bed, the adsorbent material adsorbs (e.g., selectively adsorbs) certain elements, thus allowing for their separation from the solution.

[0108] The fixed bed adsorption process may employ an adsorption column comprising a column inlet and a column outlet, wherein a bed is positioned between the inlet and the outlet such that the solution to be treated passes through the bed when traveling from inlet to outlet. The fixed bed adsorption process and column may employ at least one bed, or any number of beds, packed with an adsorbent. For example, the fixed bed adsorption process may employ at least 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 beds. It is contemplated that a series of fixed beds may be employed to adsorb target elements from a solution. For example, the solution may first pass through a first fixed bed, then pass through a second fixed bed, then pass through a third fixed bed, etc. In such embodiments, it is contemplated that the adsorption process may reachAtty. Ref. No. 0073605-000829 equilibrium layer by layer from column inlet to column outlet. However, it is further contemplated that, due to adsorption kinetics, the boundary between the loaded and unloaded adsorbent layer may not be sharp, and instead it may be a broad zone known as a mass transfer zone (MTZ). The MTZ travels through the fixed bed column during the adsorption process from the column inlet to the column outlet. The velocity of MTZ travel is much slower than water velocity, depends on the affinity of adsorbate for adsorption, and may be slower for species with higher affinity.

[0109] In some embodiments, an adsorption column may be packed with a bed of functionalized polystyrenic beads (e.g., A-Cat, A-Cat-N, A-Py). The polystyrenic beads may have any diameter, such a between 0.30 to 1.2 mm. It is contemplated that larger beads may have the advantage of better flow characteristics and less pressure complications. It is further contemplated that an adsorption column may be packed with any aminated solid support that supports amine on its surface. There can be one or more adsorbents used in the adsorption process. Any one or combination of adsorbents can be the same or different (e.g., same or different in structure, in chemistry, etc.). Amounts or concentrations of one adsorbent can be the same or different from the amount or concentration of another adsorbent. Any one or combination of adsorbents can be homogenous or heterogenous in structure.

[0110] It is contemplated that the flow rate of the solution passing through the adsorption column may be 0-1 mL / min, 0.1-0.5 mL / min, preferably about 0.15 mL / min. It is contemplated that as the flow rate decreases, adsorption of the target ions may increase.

[0111] After adsorption, the target ion (e.g., germanium) can be extracted and collected from the adsorbent via a desorption process. Desorption can be achieved by introducing the adsorbent (with the metal ions adsorbed thereto) to a mineral acid, such as a hydrochloric acid.Atty. Ref. No. 0073605-000829 For example, a mineral acid may be passed through an adsorption column and the fixed bed(s) and collected at the column outlet.

[0112] It is contemplated that the adsorbents may be regenerable. It is understood that regeneration involves removal (desorption) of compounds adsorbed onto the surface of the adsorbent such that the adsorbent may be used to adsorb further compounds. The adsorbent may be regenerated using any suitable means. For example, the adsorbent may be regenerated using an acid eluent (e.g., HCl, H2SO4, or any other suitable acid). It is contemplated that the adsorbent may undergo multiple cycles of adsorption / desorption.

[0113] It is contemplated that the adsorbents may retain greater than 70% capacity after five adsorption-desorption cycles. EXAMPLES

[0114] Example 1. This example investigates synthesized adsorbents for Ge adsorption in terms of pH dependency, Ge selectivity, adsorption isotherm, adsorption kinetics, and adsorbent reusability. This example further investigates microwave-based functionalization of catechol- based compounds onto solid supports with amine groups. The example includes polystyrenic bead and silica as the base material of the aminated solid support. The synthesized adsorbents are characterized by FTIR and CHNS analysis. Materials

[0115] Purolite A110, macroporous polystyrene crosslinked with divinylbenzene with primary amine was provided. 3,4-dihydroxybenzaladehyde (>98%, TCI), 5-nitro-3,4- dihydroxybenzaladehyde (>98%, TCI), 2,3,4-trihydroxybenzaldehyde (>98%, TCI), glacial acetic acid (>99.7% ACS grade, Thermo Scientific), NaBH4(99%, Sigma Aldrich), and anhydrous methanol (MeOH, HPLC grade, Fisher Scientific) were used for the functionalizationAtty. Ref. No. 0073605-000829 of the A110. Ge oxide (GeO299.9999% metal basis, Alfa Aesar) was used for the preparation of Ge(IV) solution for adsorption experiments. HNO3(trace metal grade, BDH Aristar Plus), HCl (trace metal grade, BDH Aristar Plus), H2SO4(trace metal grade, BDH Aristar Plus), NaOH pellets (99-100% ACS grade, Millipore Sigma), and Type I ASTM water were used to prepare acidic and basic solutions. The competitive-ion solutions were made by using single-element ICP stock solutions (Inorganic Ventures). Functionalization

[0116] To synthesize the aminated polystyrenic beads (e.g., A-Cat, A-Cat-N, and A-Py), 2 g bead (50% moisture) was suspended in 40 ml of 10% acetic acid in methanol medium with 30 mmol of 3,4-dihydroxybenzaladehyde (4 g) or 3,4-dihydroxy-5-nitrobenzaldehdye (5.4 g) or 2,3,4-trihydroxybenzaldehyde (4.6 g) and shaken for 30 min. The mix was then microwaved. Following the heating in the microwave, beads were filtered, washed with methanol for 3 hours (in 50 ml tubes, end-to-end mixing, 30 rpm), filtered, and dried overnight at 60°C.

[0117] Both catechol (A-Cat) and nitro-catechol (A-Cat-N) functionalized beads were reacted with NaBH4to reduce C=N to solve the issue of hydrolysis of C═N in an aqueous medium. The reduction step was the same as the catechol-functionalized chitosan, but the amount of NaBH4was 3.4 g (90 mmol) per 1g of the bead. For washing, the functionalized beads were rinsed with Type I water multiple times, followed by washing with 2 M HCl for 24 hours (in 50 ml tubes, end-to-end mixing, 30 rpm), followed by multiple rinsed with Type I water, and a final Type I water wash for 12 hours. The adsorbent was filtered and dried in an oven at 60°C for 24 hours.

[0118] Aminated silica-gel was also functionalized with catechol (S-Cat) and nitro- catechol (S-Cat-N) using the same reaction as FIG. 3. To synthesize adsorbents with theAtty. Ref. No. 0073605-000829 aminated silica, the amount of reactant per g of dry silica was 0.95 g 3,4- dihydroxybenzaladehyde, 1.3 g 3,4-Dihydroxy-5-nitrobenzaldehdye, and 1.05 g ,3,4- trihydroxybenzaldehyde for S-Cat, S-Cat-N, and S-Py respectively. The reduction and washing steps were the same as bead functionalization. For the reduction step 0.8 g NaBH4was used to reduce 1 g of S=Cat, S=Cat-N, and S-Py. Characterization

[0119] The polystyrenic bead A110 and the functionalized adsorbents A-Cat and A-Cat- N were characterized by Fourier transform infrared spectroscopy (FTIR) using a Vertex 70 (Bruker Optics, Germany) instrument. The scanning electron microscope (Thermo Fisher Scientific VeriosG4) was used to take images at 200,000×.

[0120] The back-titration method was used to determine the amount of amino group in A110. The amines of the A110 reacted with excess HCl for three days at room temperature (20°C-22°C) with end-to-end mixing at 30 rpm. The remaining HCl concentration was measured using a pH meter (SevenCompact S220 pH meter, InLab Expert pH electrode, Mettler Toledo), and the number of surface amino groups was calculated from the moles of acid consumed. The extent of functionalization and surface catechol density were determined by analyzing the elemental C and N using UNICUBE Elemental CHNS-O Organic Analyzer. Adsorption Studies

[0121] The adsorption experiments were conducted in polystyrene screw cap test tubes by adding 10 mL of Ge solution with 10 mg of dry adsorbent. The mixture was then placed on a rotator and mixed end-over-end at 30 rpm at room temperature. The pH of the solution was adjusted with NaOH and HCl solutions. After 24 hours of mixing, the supernatant was collected and analyzed for Ge concentration using ICP-MS (Agilent 7800) after being acidified. TheAtty. Ref. No. 0073605-000829 equilibrium adsorption capacity (qe) (Equation 1) and solid-liquid distribution coefficient (partitioning) coefficient (Kd) (Equation 2) were calculated as: ^^^^( ) � �� = ^^^^�− ^^^^�∗�(1) (2)

[0122] where Cosolution in batch adsorption (mg / L), m is the mass of the adsorbent used (g), and V is the total volume of the solution (L). Adsorption isotherm was conducted at initial pH of 3 at room temperature (20°C) to study the effect of the increase in the initial concentration of Ge and to estimate the maximum adsorption capacity. Competing-ion adsorption experiments were done at pH 2 and pH 3 in the presence of Be(II), Mg(II), Al(III), Sc(III), Mn(II), Fe(III), Cu(II), Zn(II), Ga(III), As(IV), Cd(II), La(III), Yb(III), Pb(II), Th(IV), and UO22+each with 20 mg / L concentration. The separation factor of germanium over other ions in the competing-ion adsorption is calculated as (Equation 3): SF(^^^^ ^^^^ / ^^^^)=�� (3)

[0123] Results

[0124] Functionalization and Characterization. The number of amino groups on unfunctionalized A110 was determined to be 5.9 mmol / g with back-titration. The extent of catechol and nitro-catechol functionalization on A110 and the silica gel was calculated using the elemental C / N ratio determined by the Elemental CHNS-O Organic Analyzer (Table 1). Table 1: Surface Ligand Density of Different Synthesized Adsorbents Surface ligand Adsorbent CN analysis (%) density (mmol / g)Atty. Ref. No. 0073605-000829 Ligand C N C / N grafting Amine Catechol yield (%) A110 78.1 8.1 9.7 na* 5.9 na* A-Cat 67.2 5.3 12.8 44.0 3.3 2.6 A-Cat-N 59.4 6.9 8.6 70.0 1.8 4.1 A-Py 64.3 4.8 13.3 51.4 2.9 3.0 Silica-NH27.2 2.3 3.1 na* 1.4 na* S-Cat 8.0 1.6 5.0 26.6 1.0 0.4 S-Cat-N 8.0 2.0 4.0 27.8 1.0 0.4 S-Py 10.3 1.8 5.8 37.7 0.9 0.5 *na = not applicable.

[0125] The functionalization yields with catechol were around 44% in the case of A110 and around 26.6% in the case of silica-NH2, and for nitro-catechol, the yield was 70% for A110, and 27.8% for silica-NH2. The functionalization yield of pyrogallol was 51.4% with A110 and 37.7 % with silica-NH2, higher than catechol and nitro-catechol (Table 1). The functionalization process involving reflux, as done previously with chitosan, was repeated for the A110. Additionally, in a control synthesis, no microwave heating was used, and beads were sampled just after 30 min mixing with the reactant. The use of microwave results in higher yield than both the control and the 24-hours reflux process (Table 2). Additionally, it reduces the heating time required for the reaction from 24 hours in reflux to 4 min in the microwave. Table 2: Reaction Yield of Different Synthesis Processes Involving Different Heating Step AdsorbentAtty. Ref. No. 0073605-000829 Avg. Ligand C / N grafting Ratio yield (%) A110 9.7A=C 0 min 12.5 39.9 A=C 4 min 13.8 58.6 A=C Reflux 13.0 46.7 *0 min and 4 min refer to no microwave heating and microwave heating for 4 mins, respectively. Reflux refers to the synthesis involving 24-hour reflux step.

[0126] Compared to solid support A110, in A=Cat (FIG. 4), A=Cat-N (FIG. 5), and A=Py a new vibration peak for C=N appears at 1658, 1657 cm-1,and 1628 cm-1respectively. The C=N peak disappears after the reduction of C=N to CH-NH, implying successful reduction. In the case of functionalization with nitro-catechol, the -NO2vibration peak appears at 1528 cm-1in the A=Cat-N (FIG. 5). After functionalization a new peak emerges at 1119 cm-1in FIG. 4, 1133 cm-1in FIG. 4, and 1081 cm-1in FIG. 6 referring to C-OH vibration of the grafted catechol.

[0127] The strong primary amine peak of A110 appears at 3280 cm-1and 3349 cm-1. The broad peak at 3200-3600 cm-1is due to vibration of O-H, N-H, and intramolecular hydrogen bonds. The change in the shape of broad peak 3200-3600 cm-1is due to grafting of catechol and catechol derivatives which adds more -OH groups onto the surface.

[0128] In the case of aminated silica gel as the solid-support, the C=N peak shows up at 1654 cm-1for S=Cat (FIG. 7) and 1659 cm-1for S=Cat-N (FIG. 8) and disappears after the reduction step. The vibration peak for -NO2shows up at 1528 cm-1for nitro-catechol functionalized silica (FIG. 8). Similar to the case of A110, primary amine peaks appear in theAtty. Ref. No. 0073605-000829 Silica-NH2at 3298 cm-1and 3365 cm-1. The shape of the spectra changes in the 3200-3600 cm-1in the silica functionalization as well due to the grafting of catechol and catechol derivatives.

[0129] The scanning electron microscope (SEM) images for A110, A-Cat, and A-Cat-N did not show differences at 200000× (FIG. 9). Additionally, the after Ge(IV) adsorption EDS of A-Cat (FIG. 10) and A-Cat-N (FIG. 11) showed that the C, O, N, and Ge are uniformly distributed on the surface, indicating that the adsorption sites are uniformly distributed on the adsorbent and not concentrated in a particular region.

[0130] Adsorption Studies. The functionalized silica adsorbents S-Cat, S-Cat-N, and S- Py showed lower adsorption capacity than the functionalized A110 adsorbents (Table 3). The higher capacities of A110 are due to a higher number of starting amine groups compared to aminated silica gel, thus resulting in higher functionalized ligand density on the surface after functionalization. The higher number of surface ligands resulted in higher adsorption capacity for Ge(IV). Due to the higher adsorption capacity, only functionalized A110 adsorbents were considered for further experiments. Table 3: Ge(IV) Adsorption Capacity of Different Synthesized Adsorbents Ge Surface ligand adsorption Adsorbent density capacity (mmol / g) (mg / g) A-Cat 2.6 28.1 ± 1.1 A-Cat-N 4.1 36.7 ± 1.5 A-Py 3 36.8 ± 2.4 S-Cat 0.4 7 ± 1.9Atty. Ref. No. 0073605-000829 S-Cat-N 0.4 7.8 ± 0.5 S-Py 0.5 9.7 ± 0.6

[0131] All three adsorbents showed good adsorption at all pHs from 1-11 (FIG. 12). The adsorption increased with increase in pH for A-Cat until pH 3 and remained similar at higher pHs except pH 11 at which it decreased. The adsorption remained similar for A-Py at all pH except pH 11, whereas for A-Cat-N, the adsorption increased with pH until pH 6 thereafter, it started to decrease until pH 11. The equilibrium pH after adsorption is different from the initial pH (FIG. 13). The equilibrium pH is affected by the protonation-deprotonation reaction of the surface -NH2and surface -OH groups and the Ge-catechol complexation reaction. The number of surface -NH2and surface -OH groups, as well as the pKa value of the hydroxyl group, are different for the A-Cat, A-Cat-N, and A-Py; thus, the equilibrium pH differs for these adsorbents. The solution pKa values follow 4-nitro-catechol (6.78, 10.9) < pyrogallol (9.05, 11.2) < catechol (9.45, 13), and such a trend will result in lower equilibrium pH for A-Cat-N compared to A-Cat.

[0132] The effect of initial pH on adsorption is different from the effect of initial pH on Ge-ligand complexation in free solution (FIG. 14). FIG. 14 shows the theoretical Ge-L fraction at different pH. The pKa values and Ge-ligand complexation stability constants for catechol and nitro-catechol were taken from Patel and Karamalidis, and for pyrogallol, pKa values were obtained from, and the Ge-Pyragallol complexation stability constants from Pichet and Benoit. In free-solution, the fraction of the Ge-catechol complex and Ge-pyrogallol at pH<2 is very low (FIG. 14). The catechol-functionalized adsorbent shows similar trend to the free solution complexation; however, the adsorption at pH 1 and pH 2 is high as opposed to free-solution complexation (~ 0%, FIG. 14). Similary, pyrogallol-functionalized adsorbent also showed goodAtty. Ref. No. 0073605-000829 adsorption at pH <2 and the adsorption did not decrease at pH<2. This change in complexation behavior of catechol and pyrogallol after grafting indicates that the free-solution pH dependency may not be followed exactly in the case of functionalized adsorbent. It has been reported that the pKa values of ligands change after functionalization. Thus, it is assumed that the Ge- complexation behavior of the ligands has changed after functionalization and is also affected by other parameters such as surface charged and electrical double layer as well as difference in complexation mechanism between surface ligand and the in-solution ligand. Similarly, in the case of nitro-catechol, the free-solution Ge-L3fraction is reduced to around 60% at pH 1, however in A-Cat-N, the capacity at pH 1 is 24.4 mg / g, which is around 71% of 34.3 mg / g observed at pH 6. Thus, the ligand behaves differently in the case of A-Cat-N as well; however, the difference in the behavior is not evident as in A-Cat and A-Py, where adsorption occurs at pH<2 despite no Ge-ligand complexation in free solution.

[0133] Competitive adsorption experiments with A-Cat and A-Cat-N were done in the presence of competing ions Be(II), Mg(II), Al(III), Sc(III), Mn(II), Fe(III), Cu(II), Zn(II), Ga(III), As(IV), Cd(II), La(III), Yb(III), Pb(II), Th(IV), and UO22+, each with 20 mg / L concentration (prior to precipitation of some ions such as As). Competitive adsorption was carried out with initial pH 1, pH 2, and pH 3 to avoid precipitation of ions such as Al, Fe, Ga, and Pb due to changes in the pH of the solution during adsorption. Thorium can precipitate at pH >2; thus, Th(IV) was excluded in the pH 2 and pH 3 competitive ion adsorption experiment.

[0134] The selectivity of A-Cat, A-Cat-N, and A-Py is studied with solid-liquid distribution (partitioning) coefficient (Kd) and selectivity factor. The distribution coefficient (Kd) is the ratio of the amount adsorbed to the amount remaining in the solution and depends on aqueous chemical conditions (e.g., pH, adsorbate concentration, solution / electrolyte matrix,Atty. Ref. No. 0073605-000829 temperature). The selectivity factor is the ratio of Kdfor the target element with the Kdof the competitive element.

[0135] The Ge distribution coefficients for A-Cat were 1427.7 mL / g, 4801.5 mL / g, and 7348.9 mL / g at pH 1, pH 2, and pH 3, respectively. The distribution coefficients were 3282.4 mL / g and 1823.5 mL / g at pH 1, 5270.5 mL / g and 2648.5 mL / g at pH 2, and 6431.0 mL / g and 2788.0 mL / g at pH 3, respectively for A-Cat-N and A-Py. The Kdvalues for Ge(IV) are significantly higher than other elements (FIG. 15), and thus, all three adsorbents were selective for Ge against other elements at all three pHs (FIG. 16). The lowest selectivity factors for Ge(IV) were observed to be 14.2 against Ga(III) for A-Cat at pH 3, 6.3 against Fe(III) for A-Cat-N at pH 3, and 3.6 against Fe(III) for A-Py at pH 3. The selectivity factors against other elements were higher than these minimum values, and for many competing elements, adsorption was not observed. The A-Cat was more selective for Ge(IV) than A-Cat-N and A-Py in general since SF(A-Cat) / SF(A-Cat-N) >1 as well as SF(A-Cat) / SF(A-Py) >1 (FIG. 16).

[0136] For negatively charged O-donor atom ligands, the ligand-metal stability constants (log Kaq) in solution are linearly correlated with the metal hydrolysis constants, and the slope of the linear correlations increased with the increase in ligand basicity (∑pKa). Linear free energy relationship (LFER) between the adsorption and first hydrolysis constant of the metal cations was investigated. The apparent equilibrium constant, K(ads), for the adsorption is calculated by normalizing the distribution coefficient, Kd,with the surface site density (Equation 5). The adsorbents A-Cat, A-Cat-N, and A-Py have different surface site densities that will affect the Kdvalue, hence the need for normalization. �^ ^^^^ ^^^^ ^^^^ ^^^^����∗(5)Atty. Ref. No. 0073605-000829

[0137] Where Kdis the distribution coefficient in mL / g, nsis the surface density (in mmol / g) of catechol, nitro-catechol, and pyrogallol in the A-Cat, A-Cat-N, and A-Py, respectively.

[0138] Additionally, the unit of Kadsagrees with the theoretical adsorption reaction of the form (Equation 6): ≡L (mmol / g) + Mn+(mol / L) = ≡L-Mn+(mmol / g) (6)

[0139] With the equilibrium constant for the reaction being (Equation 7): ^ ={≡L− ^^^� }{≡L}{M ^^^^+}(L / mol) (7)

[0140] Where ≡L is the surface ligand (in mmol / g), Mn+is the metal cation (in mol / L), and ≡L-Mn+is the ligand-metal association on the surface ligand (in mmol / g). The reaction assumes 1:1 stoichiometry, which is not the case for many elements. Additionally, the reaction assumes that the cation exists as a single species. The proposed K(ads) has the same limitations as Kdand thus is a function of solution condition (e.g., pH, adsorbate concentration, solution / electrolyte matrix, temperature), and thus, it applies only to the solution in which it was measured.

[0141] Linear trends were observed between log Kadsand the first hydrolysis constant of the metal cations. The log Kadsof Ge, As, and Yb were not considered since they lack experimental hydrolysis constants. Results with Kdvalues <5 mL / g were excluded in the LFERs (may have resulted from experimental uncertainty). The data from competitive ion adsorption at pH 1 was not included in LFER due to most ions not showing any adsorption.Atty. Ref. No. 0073605-000829

[0142] The intercept of the LFER between log Kadsand log K(MOH) is forced to zero. A non-zero intercept results from using units other than molarity to express concentration, or if the equilibrium constants are expressed in mole fractions. Additionally, zero intercept is a convenient feature as only one metal-specific parameter is required to estimate the adsorption equilibrium. Moreover, the different ligands can be compared based on the slope of the LFER. Physically, a value of zero for log K(MOH) means that the metal does not form a bond with -OH in free solution, and Gibb’s free energy of formation is zero; thus, it is assumed that such a hypothetical element will not get adsorbed onto the surface -OH as well, resulting in zero log Kads.

[0143] The applicability of such a correlation with catechol-functionalized chitosan has previously been shown. Similar to the catechol functionalized chitosan the high Kdand selectivity for Ge can be understood based on the LFERs. Ge(IV) has a higher affinity for hydrolysis than the elements used in LFER; thus, it will have a much higher logK(MOH) (>14) than the cations used in the LFERs, resulting in higher log Kdbased on the LFERs. The slope follows the order of A-Py>A-Cat>A-Cat-N at both pH 2 and pH 3 (FIGS. 17 and 18). The slope of the LFER between log K(metal-ligand complex) and logK(MOH) decreases with decrease in ∑Pka of the ligand in solution. The solution pKa values of these ligand follows catechol (9.45, 13) > pyrogallol (9.05, 11.2) > 4-nitro-catechol (6.78, 10.9), thus the order of slope in free solution should be catechol>pyrogallol>nitro-catechol. The order of the slope here differs from expected order of slope based on ∑Pka. This is due to the fact that the Kadsare not independent of pH. The Kadsrepresented here are a snapshot of adsorption equilibrium at a particular pH and concentration and ignore the existence of possible species (such as Fe3+, Fe(OH)2+, etc., all of them will have their own Kads). Additionally, the different elements can have different adsorptionAtty. Ref. No. 0073605-000829 mechanisms (monodentate, bidentate, or polydentate complexation), but the Kadscannot account for such differences. The true adsorption equilibrium constants can be obtained via surface complexation modeling (SCM). Additionally, using Kadsresulted in low R2value due to lack since Kadsare dependent on many factors, as previously explained. In cases where SCM was used to estimate log Kads,better fits have been obtained.

[0144] Despite the above limitations, the LFERs presented here offer an easy determination the selectivity of the adsorbent against other cations using the free solution thermodynamic properties of the cations. With the presented LFER, a log Kadsfor an element under pH 2 and pH 3 for these adsorbents can be obtained without an adsorption experiment.

[0145] The equilibrium Ge adsorption capacities of the A-Cat, A-Cat-N, and A-Py were determined using adsorption isotherm study at pH 3. The adsorption equilibrium is studied by fitting the experimental data with Langmuir isotherm model (Equation 8) and Freundlich isotherm model (Equation 9). ^^�� � ^^� ∗���� =(∗ )    ⇒  ��  =  +��(8) (9)

[0146] Themg / g, 39.14 mg / g, and 37.13 mg / g respectively, which matches the maximum experimental uptakes (FIG. 19, Table 4). Agreement between experimental data and Langmuir isotherm for adsorbents implies that the adsorption sites are homogenous, and the adsorption was monolayer. Based on the proposed chemical structure of the adsorbent, the adsorption sites are catechol ligand in A-Cat, nitro-catechol in A-Cat-N, and pyrogallol in A-Py. The sites are homogenous and are involve in complexation thus resulting in monolayer adsorption.Atty. Ref. No. 0073605-000829 Table 4: Estimated Adsorption Isotherm Parameters for Ge Adsorption onto A-Cat, A-Cat-N, and A-Py Based on Experimental Data and Langmuir and Freundlich Isotherm Fitting Exp. Langmuir isotherm Freundlich isothermMax qeQmKLKFR2n R2A-CatA-Cat-N 36.95 39.14 304.83 0.99 12.04 3.20 0.98 A-Py 35.66 37.13 386.64 0.99 12.78 3.48 0.98

[0147] The Ge(IV) adsorption capacities of A-Cat, A-Cat-N, and A-Py are the highest among other catechol-based adsorbents (Table 5). All three adsorbents have shown excellent selectivity for Ge(IV) against other elements (FIG. 16) at pH 1-3. Better selectivity was observed at pH 1, with no adsorption occurring for many competing ions. For A-Cat, the Ge(IV) selectivity factors ranged from 14.2 against Ga(III) pH 3 to no adsorption of multiple elements at all three pHs. The lowest selectivity factor for A-Cat-N was 6.3 against Fe(III) at pH 3 and 3.6 against Fe(III) for A-Py at pH 3. All three adsorbents showed better selectivity at pH 1, providing the advantage of working at lower pH since Ge is usually leached out in acidic media. Table 5: Comparison of Adsorbents Used for Ge Adsorption in Terms of Capacity and Selectivity Capacity Selectivity results Adsorbent Ge selectivity (mg / g) factors rangeAtty. Ref. No. 0073605-000829 Competing ion (lowest SFa(ion) - solution highest SF (ion)) composition 3D-α-FeOOH 55.5 (pH - - 7.5) Activated carbon 10.5 (pH 5) As, Sb, Si, Ni (pH 0 (Ni) - 100 (Sb) (AC)b10) Catechol- <2 (pH 6) Si, Zn (pH 6) 5 (Zn) - 28 (Si) formaldehyde resin Catechol- 6.08 (pH B, Te (pH 4.5) 2.1 (Te) - 74 (B) functionalized silica 4.5) di(2- 130.7 (pH B, Te, As, Se (pH 8) 5 (Te) - 12 (B) hydroxyethyl)amine 8) grafted cellulose Kelex-100 23 (pH 6.2) As, Sb, Zn, Ni (pH 28.6 (Zn) - no ads.c6.2) (Sb) DEA-EPI-CS 23.74 (pH - - 4) HEMA-IDOL-CS 73.91 (pH Cu, Fe, Zn, Mn, Pb, 0.53 (Pb, pH 1) - 4) As (pH 1-8) 4.52 (Mn, pH 8) Purolite S108 78.92 (pH Al, Fe, Cu, Zn, Ga, 0.2 (As, pH 2) - 3) As, Cd, Sb, Pb (pH 1860.8 (Cd, pH 3) 2-3)Atty. Ref. No. 0073605-000829 C-Cat 22.95 (pH Al, Fe, Cu, Zn, Ga, 6.8 (Ga, pH 3) - no 3) As, Cd, Sb, Pb (pH ads. (Cu, Zn, Ga, 2-3) As, Pb, pH 2) A-Cat 28.21 (pH Be, Mg, Al, Sc, Mn, 14.2 (Ga, pH 3) -no This 3) Fe, Cu, Zn, Ga, As, ads. (multiple Example Cd, La, Yb, Pb, Th, elements, pH 1, pH (see U 2, pH 3) FIG. 16) A-Cat-N 36.95 (pH Be, Mg, Al, Sc, Mn, 6.3 (Fe, pH 3) -no This 3) Fe, Cu, Zn, Ga, As, ads. (multiple Example Cd, La, Yb, Pb, Th, elements, pH 1, pH (see U 2, pH 3) FIG. 16) A-Py 35.61 (pH Be, Mg, Al, Sc, Mn, 3.6 (Fe, pH 3) -no This 3) Fe, Cu, Zn, Ga, As, ads. (multiple Example Cd, La, Yb, Pb, Th, elements, pH 1, pH (see U 2, pH 3) FIG. 16)aSF: Selectivity factor (^^^^ ^^^^ ^^^^ ^^^^^^^^ ^^^^). ^^^^bCatechol was used to complex Ge in solution, followed by adsorption using AC. cno ads.: adsorption of these ions was not detected (meaning high selectivity for Ge against the particular ion).

[0148] The adsorption increased with increase in time for all three adsorbents (FIG. 22). There is a pseudo-equilibrium reached for all the adsorbents at around 12 hours. Such pseudo- equilibrium can be used to design the adsorption process so that the required time for adsorption process can be reduced. However, adsorption site saturation will not be reached in such cases.Atty. Ref. No. 0073605-000829

[0149] The kinetics of adsorption mass transfer can be divided into three sequential stages: 1) external diffusion (also known as film diffusion), 2) intraparticle diffusion, and 3) adsorption on the active sites. The adsorption kinetics were investigated using pseudo-first-order (PFO) (Equation 10), pseudo-second-order model (PSO) (Equation 11). In the PFO model the driving force is the concentration gradient and is dependent on mass transfer from bulk to the surface whereas in the PSO assumes chemisorption as the rate-limiting step. �= ^^^^�( ^^^^�− ^^^^�) ⇒ ^^^^�= ^^^^�(1 − exp(− ^^^^�^^^^)) (10)

[0150] where qtequilibrium, respectively; k1(h-1), and k2(g / (mg*h))) are the rate constants for pseudo-first-order and pseudo-second-order. Non-linear least-squares curve fitting using Matlab was used for parameter estimation for the kinetic models, as non-linear methods can provide consistent and accurate estimates of parameters.

[0151] The PSO better explains the kinetics data for all three adsorbents A-Cat, A-Cat-N, and A-Py, than PFO (Table 6). The better fit of PSO indicates adsorption rate-limiting steps to be chemisorption of Ge onto surface ligand for all three adsorbents. Table 6: Estimated Adsorption Kinetics Parameters for Ge Adsorption onto A-Cat, A-Cat-N, and A-Py Based on Experimental Data and Pseudo-1stand Pseudo-2ndOrder Fitting Exp. Pseudo-first order (PFO) Pseudo-second order (PFO) Model 103*k2qe - (mg / g) (mg / g)1) (mg / g)1)Atty. Ref. No. 0073605-000829 A-Cat 27.92 26.38 0.48 0.95 28.62 23.52 0.98 A-Cat- 39.14 36.45 0.24 0.97 40.63 7.84 0.99 N A-Py 40.83 38.35 0.26 0.97 42.61 8.53 0.99

[0152] The Ge desorption was carried out with 0.5, 1, and 2 M concentrations of HCl and H2SO4 (Table 7). The higher strength of acid concentration resulted in higher desorption. The highest adsorption was achieved with 2 M H2SO4for A-Cat, and with 2 M HCl for A-Cat-N and A-Py (Table 6). Table 7: Desorption of Ge(IV) From the Adsorbents Using Different Eluents Ge desorption (%) A-Cat A-Cat-N A-Py Acid 24 48 24 48 24 48 hour hour hour hour hour hour 0.5 M HCl 54.0 77.9 23.0 31.3 44.0 51.7 1 M HCl 59.4 85.2 28.7 40.5 46.1 50.0 2 M HCl 71.1 93.7 41.7 62.9 52.6 56.3 0.5 M H2SO451.5 74.7 25.6 35.3 41.9 46.2 1 M H2SO464.3 96.9 29.8 41.8 47.1 51.8 2 M H2SO473.1 100.0 40.5 59.3 50.7 51.7

[0153] The adsorption-desorption cycles for the adsorbent were carried out using 2 M H2SO4as eluent for A-Cat, and with 2 M HCl as eluent for A-Cat-N and A-Py.Atty. Ref. No. 0073605-000829

[0154] The A-Cat showed good reusability with consistent adsorption and desorption during the multiple adsorption-desorption cycles (FIG. 23). The other two adsorbents had low desorption in the first cycle; thus, adsorption capacity was lost. After the first cycle, A-Cat-N and A-Py showed consistent adsorption and desorption (FIG. 23).

[0155] X-ray photoelectron spectroscopy (XPS) was used to gain understanding about the adsorption mechanism. In the XPS survey scan of the adsorbents before and after Ge adsorption Ge 3d peak appears confirming that Ge was adsorbed (FIGS. 24-26). The surface Ge atomic % were 0.7 ± 0.3%, 0.6 ± 0.1%, and 0.9 ± 0.2% for A-Cat, A-Cat-N, and A-Py, respectively.

[0156] The 3d peak for metallic Ge occurs at 29.6 eV binding energy, and it shifts to 33.3 eV for GeO2.The binding energy of Ge 3d peak after being adsorbed onto A-Cat, A-Cat-N, and A-Py is 32.3 eV (FIG. 27) and is shifted 1 eV lower compared to GeO2. The shift is due to nearest neighbor effects which supports the formation of Ge-O-C. Still, it is possible that Ge is bonded as Ge-O-C as well as Ge-O-H (from Ge(OH)4). The bonding environment is dependent on adsorption mechanism and stoichiometry of Ge to ligand. In the case of 1:3 stoichiometry, Ge form 6 Ge-O-C bonds.

[0157] Example 2. This example investigates the synthesized adsorbents prepared in Example 1, particularly the adsorbents with polystyrenic bead as the base material of the aminated solid support, for Ge selectivity in the presence of competing ions in fixed-bed adsorption. The synthesized adsorbents are compared with commercially available N- methylglucamine resin in recovering Ge from synthetic Zn refinery residues solution. Aqueous complexation modeling is used to understand the difference between behavior of the ligand in the solution and ligand grafted onto the solid support. Linear Free Energy Relationship (LFER)Atty. Ref. No. 0073605-000829 is used to estimate log KMLfor different ions followed by aqueous complexation modeling to simulate the selectivity of the adsorbents. Materials

[0158] The Ge(IV) solutions for adsorption experiments were prepared using Ge oxide (GeO299.9999% metal basis, Alfa Aesar). Acid and base solutions for pH adjustment were made using HCl (trace metal grade, BDH Aristar Plus), NaOH pellets (ACS grade, Millipore Sigma), and Type I ASTM water. The 2% HNO3dilution solution for ICPMS was prepared using HNO3(trace metal grade, BDH Aristar Plus) and Type I ASTM water.

[0159] The competitive-ion solutions were made by using single-element ICP stock solutions (Inorganic Ventures). Synthetic Zn refinery residue was prepared using ZnCl2(Fisher Chemical), FeOOH (Reade International), NiSO4.6H2O (Merck), CuSO4.5H2O (MP Biomedicals), CoCl2.6H2O (BTC), Ga(OH)3(Thermo Scientific), and GeO2(Alfa Aesar). The solution contained 1000 mg / L Fe(III), Cu(II), and Zn(II) each, 40 mg / L Ga(III), and 50 mg / L each of Ge(IV), Co(II), and Ni(II) with addition to 0.01 M KCl and the pH adjusted to pH 2 using HCl and NaOH. There was precipitation of Fe(III) at pH 2; thus, the initial solution was filtered 24 hours after preparation to remove any precipitates. The actual concentration of Fe(III) after precipitation was 509 mg / L. A commercial spherical resin Purolite S108, a macro-porous polystyrene cross-linked with divinylbenzene with N-methylglucamine as the functional group was used as another adsorbent to compare with the A-Cat. Adsorption Studies

[0160] The column adsorption experiments were carried out using 1 mL FPLC columns (G-Trap Biosciences) (column diameter = 0.785 cm; column depth = 2 cm), Masterflex L / S digital standard drive peristaltic pump, and polyvinylchloride (PVC) based tubing. The ratio ofAtty. Ref. No. 0073605-000829 column diameter to average bead diameter (1.2 mm) was 654.2. The bed volume was 0.97 mL and was taken as 1 mL for practical purposes.

[0161] The flow rate and inlet Ge(IV) concentration were varied to understand the effect of both parameters and model the fixed-bed adsorption process. The flow rates were 9 BVH (0.15 mL / min), 30 BVH (0.5 mL / min), and 60 BVH (1 mL / min) with an inlet concentration of 30 mg / L Ge. The inlet Ge(IV) concentrations were 10 mg / L, 30 mg / L, and 80 mg / L, with a flow rate of 30 BVH. The ionic strengths of all the solutions were fixed at 0.01 M using KCl solution. Effluent samples were collected at regular intervals and were analyzed using ICPMS.

[0162] After reaching the required number of bed-volume treatments, the adsorbed ions were desorbed using a 2 M HCl solution. In the desorption step, 2 M HCl was passed through the fixed bed at a flow rate of 30 BVH for 120 bed volumes. Multiple adsorption-desorption cycles were conducted for the adsorbents to understand the reusability. After the desorption, the bed was rinsed with deionized water for at least 10-bed volumes to avoid pH changes in the next adsorption cycle. No desorption conditions optimization was attempted.

[0163] Competing-ion fixed-bed adsorption experiments were done at pH 2 and pH 3 with 10 mg / L Ge(IV) in the presence of Be(II), Mg(II), Al(III), Sc(III), Mn(II), Fe(III), Cu(II), Zn(II), Ga(III), As(IV), Cd(II), La(III), Yb(III), and Pb(II), each with 10 mg / L concentration. Samples were collected at appropriate bed volumes and diluted using 2% (v / v) HNO3solution, and the ion concentrations were determined using Agilent 7800 ICPMS. Mass Transfer and Zone Calculations

[0164] The height and the travel velocity of the mass transfer zone can be calculated with the following equations and using the breakthrough point when C / C0= 0.05 and the exhaustion point when C / C0= 0.95:Atty. Ref. No. 0073605-000829 ^^^^ �� b= � b��1 −��� d t (12)

[0165] Where Coand C(mL / min) is the flow rate of the Ge(IV) solution, m (g) is the amount of adsorbent, tband te(min) are the breakthrough and exhaustion time, and h (cm) is the bed height. Breakthrough and Curve Modeling

[0166] The curve between C / Coand the time (or number of bed volumes, BVs) in fixed bed adsorption is referred to as the breakthrough curve (BTC). The shape of the breakthrough curve is a fundamental characteristic for understating dynamic adsorption. Many empirical equations have been proposed for modeling breakthrough curves.

[0167] Thomas model: One of the commonly used BTC models, the Thomas model, assumes the adsorption as a pseudo-second-order reaction and has the following linear form: � �T��� ln��� − 1� =�− ^^^^T^^^^�^^^^ (16)

[0168] Where kT(L / min.mg) is Thomas constant and ^^^^�(mg / g) is Thomas adsorption capacity. Thomas model is commonly used to determine the maximum adsorption capacity and the rate of adsorption and is applicable for estimating adsorption processes where external and internal diffusion resistances are small.Atty. Ref. No. 0073605-000829

[0169] Modified dose-response model (MDR): The MDR model is an empirical model developed to minimize the errors resulting from the Thomas model, particularly at higher and lower periods of the breakthrough curve. The linear forms of the model are as follows: � �� ln =�^^^^�ln� + ^^^^�ln(t) (17)

[0170] Where ay is MDR constant and ^^^^�(mg / g) is MDR adsorption capacity.

[0171] Yoon-Nelson Model: Yoon-Nelson model assumes that the rate of decrease in the probability of adsorption for each adsorbate molecule is proportional to the possibility of adsorbate adsorption and the probability of adsorbate breaking through the adsorbent. � � ln� ��= ^^^^^^^^ − τ ^^^^⇒ ln��− 1� = − ^^^^^^^^ + τ ^^^^YN(18)

[0172] Where kYN(1 / min) is a constant and ^^^^(^^^^ ^^^^ ^^^^)is half-column saturation time according to Yoon-Nelson model. Results

[0173] Preliminary Column Adsorption Experiment. All the adsorbents showed Ge adsorption in the fixed bed column adsorption (FIGS. 28 and 29, Table 8). The total amount of Ge adsorbed (q) and desorbed (qD) is calculated using Equations (19) and (20), respectively. The % desorption and maximum concentration factor (CF) that indicates the times the Ge was concentrated through the desorption process are calculated using Equations (21) and (22), respectively. �� ^^^^ =�1 − � d BV (19)Atty. Ref. No. 0073605-000829 ^^^^ ^^^^ ^^^^ ( ^^) =� ∗ 100 (21) ^^^^ ^^^^ =()��(22)Table 8: Comparison Between the Three Adsorbents in Fixed-bed Adsorption Adsorption Desorptionmass Cycle tbq qbqDDes CmaxCFAdsorbent (g) (min) (mg / g) (mg / g) (mg / g) (%) (mg / L) 1st 180 24.9 3.4 19.84 79.7 1743.1 58.1A-Cat 0.535 2nd 150 23.3 4.4 21.47 92.1 2491.5 83.11st 360 23.4 9.3 8.20 35.0 354.0 11.8A-Cat-N 0.610 2nd 175 20.2 4.7 15.27 75.7 1479.0 49.31st 60 20.1 1.7 10.16 50.5 1119.8 37.3A-Py 0.585 2nd 65 16.2 1.8 14.25 87.7 2000.1 66.7

[0174] The adsorbents did not reach the exhaustion point, and thus, there was leftover adsorption capacity in all the adsorbents at the end of the adsorption cycle. Between the first and second cycles, the behavior of the adsorbent changed. The A-Cat showed consistent behavior with slightly better adsorption in the second cycle based on higher breakthrough time; nonetheless, the total amount of Ge adsorption was almost the same.

[0175] The changes are much more evident in A-Cat-N and A-Py, where the BTC have significant shifts (FIGS. 28 and 29). These large shifts are due to low Ge desorption in the firstAtty. Ref. No. 0073605-000829 cycle. The batch desorption step for A-Cat-N and A-Py also suffered from low desorption. It is possible that there are some sites are harder to desorb, and at the end of the first adsorption- desorption cycle, these sites did not desorb and thus did not participate in the second cycle; the adsorption took place in vacant sites that were desorbed or unused in the first cycles. The difference in the sites is more evident in the desorption curve of A-Cat-N, where the first desorption cycle is different from the second cycle, indicating that in second cycle, the sites being desorbed may be different.

[0176] The Ge was concentrated multiple times compared to the inlet concentration of 30 mg / L, with the highest being 83.1 times in the second cycle of A-Cat. The desorption was >90% within 130-140 min for A-Cat (>50% desorption in the first 10 min). A-Cat-N and A-Py showed similarly high initial desorption, but as noted earlier, the total Ge desorption was much lower than A-Cat.

[0177] Ge(IV) Selectivity Experiments. Although single-element Ge adsorption provides the basic understanding, it is necessary to investigate the selectivity for Ge against other elements and the purity of Ge in desorbed eluent since the source of Ge contains competing ions.

[0178] In batch adsorption, all three adsorbents were selective for Ge(IV). The adsorbents were also selective for Ge in column adsorption (FIGS. 30-42). The columns were run for 150 BV (i.e., 150 mL) at pH 2 and pH 3 with 10 mg / L as the inlet concertation of each element (consisting of Be(II), Mg(II), Al(III), Sc(III), Mn(II), Fe(III), Cu(II), Zn(II), Ga(III), As(IV), Cd(II), La(III), Yb(III), Pb(II) and Ge(IV)). The amount of A-Cat, A-Cat-N, and A-Py in the column for the selectivity experiment were 550 mg, 590 mg, and 585 mg, respectively. The amount of different ions adsorbed (or desorbed in the desorption step) (qM) was calculated in mg M / g adsorbent mass, and the ratio of the qGeto qMrepresents the selectivity for Ge (FIG. 42).Atty. Ref. No. 0073605-000829

[0179] Between pH 2 and pH 3 inflow solutions, better Ge(IV) selectivity was observed for A-Cat at pH 2, while better Ge(IV) adsorption (in terms of higher breakthrough time) was observed at pH 3 (FIGS. 30-33). The total percent Ge adsorption (out of 1.5 mg Ge mass influx) was 98.2 at pH 3 compared to 92.7 at pH 2. The higher selectivity at pH 2 but a lower amount of Ge(IV) adsorption in fixed-bed adsorption is analogous to the batch multi-element adsorption experiments.

[0180] The Ge(IV) adsorption selectivity (qGe / qM,M: element) was highest against La(III) at pH 2 and Mg(II) at pH 3, with values of 92.3 and 141.8, respectively (FIG. 42). The lowest selectivity was observed against As(V) at pH 2 and Ga(III) at pH 3, with values of 6.7 and 2.3, respectively.

[0181] The desorption was carried out with 2 M HCl, and samples were initially collected for each BV for the first 10 BV. The qGe / qMin the desorption step represents the ratio of the total amount desorbed. The selectivity in the desorption step is more critical than the adsorption step since it results in a more purified Ge solution. For the desorption step of the pH 3 case, in the first BV, more Ga was observed than Ge (FIG. 33), likely due to the faster desorption of Ga compared to Ge. For the pH 2 case, the desorbed eluent has more Ge. More purified Ge effluent can be produced by desorbing weakly adsorbed species using a desorption step involving water or weaker acids followed by desorption of Ge(IV) using 2 M HCl.

[0182] The A-Cat-N and A-Py showed higher Ge(IV) adsorption with 99.1% and 95% for A-Cat-N and 99.0% and 97.6% for A-Py at pH 2 and pH 3, respectively; however, the Ge(IV) selectivity was lower than A-Cat (FIG. 42). These results align with the batch adsorption results where A-Cat-N and A-Py have higher adsorption capacities but lower selectivity than A- Cat.Atty. Ref. No. 0073605-000829

[0183] However, A-Cat-N and A-Py exhibited very low desorption of Ge(IV) (FIGS. 34- 42). Due to the low Ge(IV) desorption and higher desorption of competing ions, the desorption step was not selective for Ge(IV). This issue can be solved using different lixiviants, or these two adsorbents can be used to selectively separate Ga(III), another critical element.

[0184] Effect of Varied Flow Rate and Varied Concentration. The flow rate of the solution through the column affects the adsorption process. Slower flow results in higher contact time between the liquid and the solid phase, leading to better retention of the adsorbate. The decrease in breakthrough bed volume (C / Co= 0.05) with an increase in flow rate is evident in FIG. 43, which illustrates breakthrough curves for various flow rates under study. The lowest flow rate (9 BVH, 0.15 mL / min) led to the longest breakthrough bed volume and highest adsorption amount at breakthrough (Table 9). Higher breakthrough bed volumes can be achieved with even lower flow rates. Table 9: Column Adsorption Parameters, Mass Transfer Zone Parameters, and Empirical Model Fitting Results for Different Operating Conditions Operating Conditions 1 2 3 4 5 Adsorption Q mL / min 0.5 0.15 1 0.5 0.5 parameters Co mg / L 30 30 30 80 10 m g 0.535 0.535 0.5 0.58 0.52 h cm 2.00 2.00 2.00 2.00 2.00 A cm20.48 0.48 0.48 0.48 0.48 u cm / min 1.03 0.31 2.07 1.03 1.03 q mg / g 24.5 29.2 22.6 60.4 9.4Atty. Ref. No. 0073605-000829 BV (at C / Co = 0.05) 90 390 45 84 390 qbmg / g 4.96 21.64 2.62 16.31 7.05 Ads (%) 81.0 96.5 74.5 74.8 92.8 Mass HMTZcm 1.64 0.56 1.79 1.48 0.73 Transfer vMTZcm / min (x103) 2.03 0.56 4.78 2.19 1.74 Zone Thomas kTL / mg.min (x105) 15.0 10.8 27.2 8.0 47.3 model qTmg / g 27.6 30.0 25.1 62.8 11.1 R2 0.80 0.93 0.91 0.84 0.89MDR model ay 1.5 4.0 1.3 2.0 1.6qYmg / g 41.3 46.7 28.3 80.6 22.3 BV (at C / Co = 0.95) 3327 4782 1226 1739 4615 R2 0.99 0.98 0.96 1.00 0.99Yoon-Nelson kyn 1 / min 4.5 3.2 8.2 6.4 4.7 model τ min 986 3571 419 911 1150 R20.80 0.93 0.91 0.84 0.89

[0185] The mass transfer zone parameters and BTC models provide insight into the adsorption process and are used to compare different operating conditions (Table 9). The height of the mass transfer zone (HMTZ) was calculated using the adsorption capacity from the Thomas model as the capacity at the exhaustion point since the exhaustion point was not reached. The HMTZfollowed the trend case 2 (9 BVH, 30 mg / L Ge(IV)) < case 5 (30 BVH, 10 mg / L Ge(IV)) < case 4 (30 BVH, 80 mg / L Ge(IV)) < case 1 (30 BVH, 30 mg / L Ge(IV)) < case 3 (9 BVH, 30Atty. Ref. No. 0073605-000829 mg / L Ge(IV)). The height of the mass transfer zone represents the spread of the zone where adsorption is occurring. There was a trend of increase in HMTZwith the increase of flow rate, but there was no clear trend in the case of increasing concentration. Similar trends were seen with other adsorbents. The vMTZalso increased with increased flow velocity since MTZ moves faster with higher flow rates.

[0186] Three different BTC models, the Thomas model, MDR Model, and the Yoon- Nelson model, were used to model the breakthrough curve (Table 10). The MDR model showed a much better fit than the Thomas and Yoon-Nelson model, both of which showed the same R2. Identical R2is because the regressions done for both the Thomas and Yoon-Nelson models are between the same dependent and independent variables (i.e., ln���� and time, t). Nonetheless, due to their distinct underlying assumptions, these models provide differing insights into the system.

[0187] The Thomas adsorption capacities were consistent at different flow rates and matched well with the Langmuir adsorption capacity of A-Cat (29.76 mg / g). The adsorbent capacities were affected by inlet Ge(IV) concentration and increased with an increase in inlet concentration. The Thomas kinetic parameter, kT, increased with an increase in flow rate but decreased with an increase in concentration. In the case of MDR model, the capacities were not consistent, and aydid not show the same trend as the kinetic parameter of the Thomas model. The half-bed saturation time ( ^^^^) according to the Yoon-Nelson model, increased with a decrease in flow rate (Table 9) The increase in ^^^^ was not proportional to the decrease in the flow rate, indicating that the increase is not just due to lower mass influx. The increase in ^^^^ with the decrease in Cois expected since lower Cowould result in more time required for bed saturation. This trend resembles other adsorbents investigated in fixed-bed adsorption. These mathematicalAtty. Ref. No. 0073605-000829 representations of the small-scale dynamic fixed-bed column system offer the potential to reduce design costs and facilitate the transition of laboratory setups to industrial applications.

[0188] Column Regeneration and Reuse. The A-Cat adsorbent was subjected to multiple cycles of adsorption and desorption steps to investigate the regeneration and reuse of the adsorbent (FIGS. 45 and 46). The adsorbent showed good reusability in terms of capacity, as it retained >70% of the initial capacity after five cycles (Table 10). One of the contributors to the loss of adsorption capacity during subsequent adsorption cycle is incomplete desorption in initial cycles (Table 10). The breakthrough time decreased with the increasing number of cycles after the first two cycles, likely due to the loss of some adsorption sites. The average breakthrough time was 161 mins, and the average capacity was 23.5 mg / g (Table 10). The HMTZremained similar in each cycle, indicating similar mass transfer characteristics. Table 10: Different Adsorption and Desorption Parameters for Fixed-bed Column Adsorption- Desorption Cycles of A-Cat Cycles 1 2 3 4 5 Avg. ± S.D. Adsorp. tbmin 180 180 133 155 158 161.2 ± 19.7 qbmg / g 5.0 5.0 3.7 4.3 4.4 4.4 ± 0.5 q mg / g 24.5 24.3 20.9 18.1 17.2 21 ± 3.4 qTmg / g 27.6 27.4 22.9 20.2 19.5 23.5 ± 3.9 HMTZcm 1.48 1.47 1.51 1.42 1.40 1.46 ± 0.05 Ads. (%) 81.0 80.4 69.2 59.7 56.9 69.4 ± 11.2 Desorp. qDmg / g 19.8 21.5 18.9 16.3 16.8 18.7 ± 2.1 Des. (%) 80.9 88.2 90.2 90.2 97.8 89.4 ± 6Atty. Ref. No. 0073605-000829 1912.3 ± Cmaxmg / L 1743.1 2491.5 1844.4 1844.1 1638.2 334.9 CF 58.1 83.1 61.5 61.5 54.6 63.7 ± 11.2

[0189] Desorption was done using 2 M HCl at a flow rate of 30 BVH (i.e., 0.5 mL / min). Desorption steps showed 50% of desorption occurring in 10 minutes and >90% in 80 minutes. The total desorption increased with subsequent cycles and reached 100% in the fifth cycle. This increase is due to lower Ge(IV) in later cycles and possible desorption of Ge(IV) not desorbed in the previous cycle. The concentration factor remained high, indicating that concentrated effluent of Ge(IV) was obtained in each cycle. The average concentration factor was 63.7.

[0190] Synthetic Zn Refinery Residue Sample. A synthetic Zn refinery residue sample containing both Ge(IV) and Ga(III) was prepared based on known compositions. The composition was 1000 mg / L Fe(II), Cu(II), and Zn(II) each, 40 mg / L Ga(III), and 50 mg / L each of Ge(IV), Co(II), and Ni(II) with addition of 0.01 M KCl. The pH was adjusted to pH 2 using NaOH. There was precipitation of Fe(III) at pH 2; thus, the initial solution was filtered 24 hours after preparation to remove any precipitates. The actual concentration of Fe(III) after precipitation was 509 mg / L. For the adsorption step, the flow rate was kept at 9 BVH (= 0.15 mL / min), while for the desorption step, the flow rate of 30 BVH (= 0.5 mL / min) was used to reduce the overall downtime of the adsorbent column.

[0191] The desorption step was done in three steps. First, for 5 BV, Type I ASTM water was used as the eluent to remove very weakly adsorbed elements, followed by 0.1 M HCl as eluent for next 5 BV to remove elements with lower affinity for catechol. Finally, 2 M HCl wasAtty. Ref. No. 0073605-000829 used to desorb the rest of the elements. Such a scheme can be utilized to generate purer streams of target elements during the desorption step.

[0192] A-Cat was selective for Ge(IV) in the synthetic Zn refinery residue solution (FIG. 47). The breakthrough volume (BV at C / Co = 0.05) was around 50 BV, and the exhaustion point (C / Co = 0.95) was around 280 BV. The number of BV treated can be increased by increasing the height of the column. Desorption with water resulted in the desorption of Zn, Cu, and Fe, all of which were in high concentration in the inlet flow (FIG. 48). Desorption with 0.1 M HCl desorbed Fe and Ge primarily, with Fe desorption being more prominent. In desorption with 2 M HCl, binary solutions of Ge and Fe were obtained with Ge concentration being more than Fe. An average Ge selectivity factor of 2.2 against Fe was obtained during 2 M HCl desorption despite the initial solution having Fe more than 10 times of Ge (representing Ge concentration factor / Fe concentration factor ≈ 22). Nonetheless, Fe can be removed from the initial solution via a pretreatment step, where the pH is adjusted to pH 3-4 and Fe is precipitated out.

[0193] The commercial resin S108 showed some selectivity for Ge, but it also adsorbed all other elements and was saturated very quickly (at around 30 BV) (FIG. 49). The amount of Fe, Cu, and Zn adsorbed in mg / g was greater than the amount of Ge adsorbed. The same can be seen from the desorption curve (FIG. 50). Due to low selectivity, adsorption of competing ions, and low saturation bed volume, the use of S108 in industrial applications to recover Ge is severely limited.

[0194] During the desorption step of S108, similar to A-Cat, the water desorption step resulted in the desorption of Fe, Cu, and Zn, and some Ge desorption was obtained with 0.1 M HCl; however, with 2 M HCl, concentrated solutions of Fe were obtained. Using a pretreatmentAtty. Ref. No. 0073605-000829 step to remove Fe could work, but the issue of low Ge selectivity and competing ion adsorption will remain, limiting the applicability of S108.

[0195] Adsorption Mechanism. In solution, Ge(IV) only forms ML3complex with catechol and 1:1 and 1:2 Ge(IV): Catechol complexes have not been observed. However, Ge(IV) adsorption onto A-Cat was observed at pH < 3. The formation of only ML3complexes on the surface of A-Cat cannot explain the Ge(IV) adsorption at pH <3. Hence, the formation of pseudo-species ML and ML2during adsorption was assumed. The adsorption % was used to determine pseudo–log K (ML) and pseudo-log K (ML2) for Ge with catechol, 4-nitro-catechol, and pyrogallol using PHQREEC-PEST optimization (supplementary information). The pH adsorption data was taken from Patel et al. and converted to % of Ge complexed with ligand (FIGS. 51-52). The following reactions for Ge-ligand were proposed (Equations (23) and (24)) and considered (Equation (25)), with log K1and log K2being unknown and log K3being known: Ge(OH)4+ H2L = Ge(OH)2L + 2H2O log K1(23) Ge(OH)4+ 2H2L = Ge(OH)L2- + H++ 2H2O log K2(24) Ge(OH)4+ 3H2L = GeL32-+ 2H++ 4H2O log K3(25)

[0196] Based on the modeling, the surface-grafted catechol formed 1:1, 1:2, and 1:3 complexes with Ge(IV) (FIG. 53), whereas surface-grafted nitro-catechol and pyrogallol formed 1:1 and 1:3 complexes with 1:3 being dominant. The 1:1 Ge-catechol surface complex would be easier to desorb than the 1:3 Ge-catechol, which is much more complex and rigid in structure and would be difficult to attack with H+. At pH 3, the 1:1 complex is dominant for A-Cat, which can explain the high desorption, while for A-Cat-N, 1:3 is dominant at pH 3, resulting in difficulty in desorption. However, this simplification of the adsorption process fails to explain the lower desorption in A-Py since, as per the modeling, grafted pyrogallol has the 1:1 complex at pH 3.Atty. Ref. No. 0073605-000829

[0197] The selectivity of the adsorbents for Ge(IV) was analyzed using aqueous metal- ligand complexation modeling using log KML. However, limited thermodynamic data is available especially for nitro-catechol and pyrogallol. Thus, in case of a lack of log KML, they were estimated using LFERs for elements and ligands of interest.

[0198] LFERs are based on the linear correlation between logarithms of equilibrium constants of one series of reactions with logarithms of equilibrium constants of related series of reactions and have been used to estimate stability constants. For negatively charged O-donor atom ligands, the ligand-metal stability constants (log KML) are shown to be linearly correlated with ligand basicity (pKa) of different monodentate and bidentate ligands. These stability constants are also linearly correlated with the metal hydrolysis constants, and the slope of the linear correlations increased with the increase in ligand basicity (∑pKa).

[0199] Carbonaro et al. developed LFER to estimate the stability constant of bidentate complexes for 20 metal ions: ^^^^ ^^^^ ^^^^ ^^^^= ^^^^∗ ^^^^�∗ ( ^^^^ ^^^^ ^^^^ ^^^^+ ^^^^ ^^^^ ^^^^ ^^^^) (26)

[0200] Where − ^^^^is the slope correction factor, ^^^^�is Irving-Rossotti slope, estimated from regression of log KMLversus log KHL,and log KHL1and log KHL2are the basicity constants of the bidentate ligand and are equivalent to the pKaof the ligand. The term ^^^^accounts for the additional free energy for metal-ligand complexation due to the chelating effect. The subscript “OO” was used to designate that the ring is formed by two oxygen donor atoms. The value of ^^^^�is available for 20 metal ions.

[0201] The ^^^^�values are well correlated with the log KMOH(FIG. 54). Thus, the correlation can be used to estimate the ^^^^�values beyond the 20 metal ions for which these values have been determined. A universal value of 1.44 was reported for ^^^^, however, for the catechol ^^^^value of 1.44 overestimated the log KML, and thus, the new universal value of ^^^^wasAtty. Ref. No. 0073605-000829 estimated using available logKMLvalues for catechol and nitro-catechol. To estimate ^^^^slope between the available logKML and ^^^^�values (estimated from logK(MOH) using FIG. 54) multiplied by ∑logKHLwas determined (FIG. 55). The value of ^^^^was found to be 1.15, which resulted in more agreement with actual log KMLvalues (FIG. 56).

[0202] The log KMLvalues were estimated for elements for which the data does not exist (presented in FIG. 57), and aqueous complexation modeling was done at pH 1-pH 4. Beyond pH 4, the fraction of ML2can become significant, for which the LFERs cannot provide log KML. Additionally, at pH>4, metal ions can precipitate, and precipitation kinetics can come into play. The amounts of ligands were kept as 3 mmol / L, 2.5 mmol / L, and 3.7 mmol / L for catechol, nitro- catechol, and pyrogallol, respectively. The complexation model was compared with actual adsorption complexation data.

[0203] The metal-ligand complexation for catechol clearly shows selectivity for the Ge (FIG. 58).

[0204] The aqueous complexation modeling for nitro-catechol and pyrogallol was mostly done using LFER estimated log KML, since limited data is available for these two ligands. In the case of nitro-catechol, the complexation curve of Fe is not in agreement with the batch adsorption results as it shows higher complexation for Fe(III) than Ge(IV) at pH 1 (FIG. 59). Interestingly, using log KMLfor Fe(III) estimated from LFER results in a complexation model that agrees with the batch and column study (FIG. 61). The complexation modeling of pyrogallol also agrees well with the high Ge(IV) selectivity obtained (FIG. 60).

[0205] The Ge(IV) complexation modeling follows the same trend as the actual data in general, but the complexation model underestimated the complexation extent of Ge at pH 2 and pH 3 for catechol and pH 1 and pH 2 for nitro-catechol. The complexation of other ions was alsoAtty. Ref. No. 0073605-000829 underestimated from the complexation model compared to actual data. This discrepancy must be due to the difference in log KMLbetween the solution and grafted surface. It is possible that the surface log KMLis higher than the solution counterpart. This can result in a higher extent of complexation in the surface and can also explain the complexation of competing ions during adsorption which is not being reflected in the aqueous complexation modeling. The more agreement between data and model for Ge is due to the use of optimized pseudo–log K (ML) and pseudo-log K (ML2). Nevertheless, the use of LFER to estimate unknown log KML,followed by aqueous complexation modeling, can provide a preliminary understanding of selectivity and can be used to select ligands.

[0206] It should be understood that modifications to the embodiments disclosed herein can be made to meet a particular set of design criteria. For instance, the number of or configuration of components or parameters may be used to meet a particular objective.

[0207] It will be apparent to those skilled in the art that numerous modifications and variations of the described examples and embodiments are possible in light of the above teachings of the disclosure. The disclosed examples and embodiments are presented for purposes of illustration only. Other alternative embodiments may include some or all of the features of the various embodiments disclosed herein. For instance, it is contemplated that a particular feature described, either individually or as part of an embodiment, can be combined with other individually described features, or parts of other embodiments. The elements and acts of the various embodiments described herein can therefore be combined to provide further embodiments.

[0208] It is the intent to cover all such modifications and alternative embodiments as may come within the true scope of this invention, which is to be given the full breadth thereof.Atty. Ref. No. 0073605-000829 Additionally, the disclosure of a range of values is a disclosure of every numerical value within that range, including the end points. Thus, while certain exemplary embodiments of the apparatus and process and / or utilization and methods of making and using the same have been discussed and illustrated herein, it is to be distinctly understood that the invention is not limited thereto but may be otherwise variously embodied and practiced within the scope of the following claims.

Claims

Atty. Ref. No. 0073605-000829 What is claimed is:

1. An adsorbent, comprising: an aminated solid support; and at least one catechol-based compound, wherein the at least one catechol-based compound is selected from the group consisting of catechol, nitro-catechol, pyrogallol, and mixtures thereof.

2. The adsorbent of claim 1, wherein the adsorbent is synthesized by microwaving the aminated solid support and the at least one catechol-based compound.

3. The adsorbent of claim 1, wherein the aminated solid support comprises a base material, wherein the base material is polystyrenic beads or silica.

4. The adsorbent of claim 1, wherein the aminated solid support comprises surface amines and wherein the ratio of surface amines to catechol-based compound is between 4:1 and 6:

1.

5. The adsorbent of claim 1, wherein the adsorbent is selective for germanium in the presence of competing ions.

6. The adsorbent of claim 4, wherein the adsorbent has a selectivity factor for germanium greater than 3.6.Atty. Ref. No. 0073605-000829 7. A method of synthesizing an adsorbent, comprising: selecting an amount of an aminated solid support and an amount of at least one catechol- based compound; mixing the selected amounts of aminated solid support and at least one catechol-based compound; and microwaving the mixture to form a microwaved product.

8. The method of claim 7, further comprising: filtering the microwaved product; and / or washing the microwaved product; and / or drying the microwaved product.

9. The method of claim 7, further comprising: reacting the microwaved product with sodium borohydride.

10. The method of claim 7, wherein the aminated solid support is provided in an aqueous medium prior to mixing the selected amounts of aminated solid support and at least one catechol- based compound.

11. A method of using an adsorbent to capture germanium from a source, the adsorbent comprising: an aminated solid support; andAtty. Ref. No. 0073605-000829 at least one catechol-based compound, wherein the at least one catechol-based compound is selected from the group consisting of catechol, nitro-catechol, pyrogallol, and mixtures thereof, the method comprising: exposing the adsorbent to a source comprising germanium; and adsorbing germanium from the source.

12. The method of claim 11, further comprising: releasing the adsorbed germanium with at least one mineral acid.

13. The method of claim 11, wherein the aminated solid support comprises a base material, wherein the base material polystyrenic beads or silica.

14. The method of claim 11, wherein the adsorbent is selective for germanium in the presence of competing ions.

15. The method of claim 14, wherein the competing ions are selected from the group consisting of Be(II), Mg(II), Al(III), Sc(III), Mn(II), Fe(III), Cu(II), Zn(II), Ga(III), As(IV), Cd(II), La(III), Yb(III), Pb(II), Th(IV), UO22+, and mixtures thereof.

16. The method of claim 14, wherein the adsorbent has a selectivity factor for germanium greater than 3.6.Atty. Ref. No. 0073605-000829 17. The method of claim 11, wherein the source is selected from the group consisting of acid mine drainage, mine tailings leachate, zinc refineries residue, and leached solutions.

18. The method of claim 17, wherein the leached solutions are selected from the group consisting of coal fly ash, fiber optics, and electronics.

19. A method of using an adsorbent to capture germanium from a source, the adsorbent comprising: an aminated solid support; and at least one catechol-based compound, wherein the at least one catechol-based compound is selected from the group consisting of catechol, nitro-catechol, pyrogallol, and mixtures thereof, the method comprising: providing at least one fixed bed comprising the adsorbent; passing the source through the at least one fixed bed; and adsorbing germanium from the source.

20. The method of claim 19, wherein the method comprises providing a series of fixed beds comprising the adsorbent.

21. The method of claim 19, wherein the at least one fixed bed is provided in an adsorption column.Atty. Ref. No. 0073605-000829 22. The method of claim 19, wherein the source is passed through the at least one fixed bed at a flow rate less than 1 mL / min.

23. The method of claim 19, further comprising: passing a mineral acid through the at least one fixed bed to release the adsorbed germanium.

24. The method of claim 23, wherein the adsorbent retains greater than 70% capacity after five adsorption-desorption cycles.

Images