A Hybrid {Silk@Zirconium MOF} Material as Highly Efficient AsIII-sponge

Exposure of humans to Arsenic from groundwater drinking sources is an acute global public health problem, entailing the urgent need for highly efficient/low-cost Arsenite (AsIII) up-taking materials. Herein we present an innovative hybrid-material, ZrMOF@SFd operating like an “AsIII-sponge” with unprecedented efficiency of 1800 mg AsIII gr−1. ZrMOF@SFd consists of a neutral Zirconium Metal-Organic Framework [ZrMOF] covalently grafted on a natural silk-fiber (SFd). ZrMOF itself exhibits AsIII adsorption of 2200 mg gr−1, which supersedes any -so far- known AsΙΙΙ-sorbent. Using XPS, FTIR, BET-porosimetry data, together with theoretical Surface-Complexation-Modeling (SCM), we show that the high-AsΙΙΙ-uptake is due to a sequence of two phenomena:[i] at low AsIII-concentrations, surface-complexation of H3AsO3 results in AsIII-coated voids of ZrMOF, [ii] at increased AsIII-concentrations, the AsIII-coated voids of ZrMOF are filled-up by H3AsO3 via a partitioning-like mechanism. In a more general context, the present research exemplifies a mind-changing concept, i.e. that a “partitioning-like” mechanism can be operating for adsorption of metalloids, such as H3AsO3, by metal oxide materials. So far, such a mechanism has been conceptualized only for the uptake of non-polar organics by natural organic matter or synthetic polymers.

urban water bodies. So far, most of the available materials have sorption capacities of the range 60-150 mg g −1 19-21 , while less than five materials achieving sorption capacities > 300 mg g −1 22,23 . In one report in 2014 23 , the highest -so far-As III uptake reported was 320 mg gr −1 by a hybrid consisting of Fe 2 O 3 nanoparticles dispersed on a macroporous silica. Our systematic efforts during the last decade led to a series of nanomaterials with promising performances, e.g. a mesoporous spinel CoFe 2 O 4 24 with an uptake of 252 mg As III g −1 , magnetic carbon nanocages 22 with a sorption capacity of 264 mg As III g −1 and MIL-100(Fe) 16 showing uptake of 120 mg As III g −1 .
Regarding the physicochemical As III -uptake mechanism, so far, in all well-understood cases, the underlying theoretical mechanism is that originally developed by Goldberg 25,26 and Manning 7,10 , which entails that: [i] As III -uptake by solid materials is determined by surface complexation of the As III species.
[ii] at pH range 5-8, i.e. typical for natural waters, the dominant species is the neutral form H 3 AsO 3 27,28 . Thus -so far -the strategy by all research groups, including us 16,22,29 , was to maximize the number and accessibility of surface sites based on diligent preparation protocols. In this way, it has been achieved a max As III -uptake capacity 320 mg gr −1 by γ-Fe 2 O 3 nanoparticles encapsulated in a macroporous Silica 23 . Recent data show that certain carbon based materials i.e. graphene-based 30 or more innovative graphydine [31][32][33] have a promising potential for adsorbing heavy metals, metalloids and other pollutants from water. In addition, some of the metal-loaded materials can have enhanced catalytic fucntionalities [34][35][36][37] .
Within this frame of thinking, aiming at maximization of the surface sites, it becomes obvious that for any material, the theoretical upper limit would be determined by the site-density and the surface area: Using this expression, a material whose maximum-As III uptake capacity is determined by surface-adsorption has an upper theoretical limit, which under ideal conditions, would be determined by its specific surface area (SSA) and the number of surface sites (Ns). In real systems, this maximum-uptake would be further limited by the binding constant of As III -on the surface-sites. On the other hand, in natural systems, Soil Organic Matter is known to be able to sequester non-polar organics via a partitioning-sequestration mechanism 38 . This phenomenon is based on the fundamental concept of the partitioning of a non-polar organic between a polar and a non-polar solvent, e.g. for example, the partitioning of phenol on [octanol: water] 39,40 . Within a technological context, when partitioning is operating, a significant mass of sorbent can be transferred in the apolar matrix, thus resulting in cost-effective molecular-separation approaches 41,42 . So far, however, such a profitable concept has not been applied to attack problems such as As III -remediation. Herein we introduce a mind-changing approach by developing a Zirconium Metal-Organic Framework [[Zr 6 O 4 (OH) 4 (NH 2 BDC) 6 ] (NH 2 -BDC 2− = 2-amino-terephthalat e) [herein codenamed ZrMOF] which is able to perform a partitioning-like As III -uptake thanks to the microenvironment of its pores. As we show, this approach allows an unprecedented As III -uptake efficiency of > 2000 mg As III per gram of ZrMOF.
Metal-organic frameworks (MOFs), which are crystalline porous materials are constituted by metal ions or metal clusters, and polytopic organic ligands, have emerged as a new class of sorbents with a promise for various remediation processes [43][44][45] . MOFs combine extremely high surface areas, well-defined pores, and a variety of functional groups. Furthermore, several MOFs show remarkable thermal (up to 400-500 °C) and chemical stability e.g. high resistance to acid or base 46,47 . Recently, several MOFs have been investigated as arsenic sorbents, mostly concerning investigations of forms of ionic-As V sorption [48][49][50] . Only a few reports exist on the capture of As III by MOFs 40,43-45 . These sorbents, however, demonstrated only moderate sorption capacities (<150 mg As III g −1 ) 51 .
Apart from the challenge of As III -uptake efficiency, the efficient large-scale utilization of As III -uptaking materials 52-54 , requires their post-synthesis engineering to be usable for large-water body cleaning. The technology of grafting of the functional material on a macroscopic surface allows scale-up handling and usage. Herein we have used woven silk fibers [SF d ] as a scaffold for grafting of the ZrMOF material. The so-obtained hybrid material ZrMOF@SF d retains the very high As III sorption capacity of ZrMOF, i.e. reaching 1800 mg As III per g −1 of material.
To understand the unprecedented high sorption efficiency of ZrMOF and ZrMOF@SF d, we have carried out a detailed study ofthe As III -sorption mechanism in conjunction with the dynamics of pore-filling and surface complexation. In a more general context, the present research exemplifies for the first time that a "partitioning-like" mechanism to be operating in for adsorption of metalloids, i.e., H 3 AsO 3 by metal oxide materials, so far conceptualized only for synthetic polymers & natural organic matter (NOM) 55 used to uptake apolar organics 56,57 .

Results and Discussion
Field Emission-Scanning electronic microscopy (FE-SEM) images showed that ZrMOF is composed of aggregated polyhedral-shape nanoparticles with size ~100-300 nm (Figure 1a1,a2). No obvious changes in shape and size of particles are observed for the material after the As III sorption (Fig. 1b1,b2). The SEM micrograph for the silk fiber (SF d ) (Fig. 1c), shows the surface morphology of well-defined fibers of natural-silk. After covalent grafting of ZrMOF on the SF d fibers, we obtain well-dispersed ZrMOF particles on the silk-fibers (Fig. 1d). Adsorption of As III onto ZrMOF@SF d does not alter the particle morphology (Fig. 1e). Thermogravimetric (TGA) analysis, Fig. S1 in Supporting Information, shows that the ZrMOF@SF d hybrid contains 5.7% w:w of ZrMOF.
For the engineering of the ZrMOF grafting on the silk fiber, we have used natural silk tissue, which we have degummed according to established procedures (see Supporting Information, photos in Scheme S1) 58 . The degumming method of Gulrajani 58 , resulted in high-quality silk-fiber, as evidenced by the SEM micrography, see Fig. 1(c), as well as the XRD pattern, Fig. 2A, which reveals retention of the fibers' order and physical integrity in the structure of the degummed silk (SF d ). After grafting of the ZrMOF@SF d, the SEM data (Fig. 1d) shows a good dispersion of the ZrMOF particles on the silk fibers. XRD data for the ZrMOF@SF d hybrid, Fig. 2a(red) show the characteristic reflection at 7.3° and 8.5° originating from the ZrMOF particles grafted on the silk. Notice that, upon grafting, the crystallinity of the SF d is distorted, i.e. see the loss of the sharp peaks at 15°-17° in Fig. 2a. This result is due to molecular covalent grafting of the ZrMOF-silane on SF d (see also Scheme S2 in Supporting Information) TGA analysis, Fig. 2c, revealed a significant weight loss (~17.2%) in the temperature range 25 to 102 °C attributed to the release of solvent molecules, mostly MeOH. Then, there is a continuous weight loss (~7.9%) (with no discrete steps) till 302 °C, followed by an abrupt weight loss (38.8%) ending at 540 °C. Finally, there is a small weight loss (~2%) from 615 to 715 °C (Fig. 2c). The TGA residue is solid ZrO 2, as confirmed by XRD data. For 100 g of ZrMOF 33.7 g of ZrO 2 was obtained after calcination, which corresponds to ~24.9% Zr w:w. Based on this % Zr content found from TGA, the suggested formula for ZrMOF is [Zr 6 O 4 (OH) 4 (NH 2 -BDC) 6   the change from 275 °C to 400 °C is associated with the breakdown of side-chain groups of amino acid residues as well as the cleavage of peptide bonds of silk fiber, and at T > 400 °C it is attributed to fibroins' degradation 59 . The TG% curves of SF and SF d exhibit a total weight loss of 98.4% and 98.9% respectively, in the range 20-700 °C with a broad exothermic peak at 400-600 °C assigned to the fibroins' degradation. The DTA curve of SF d also shows an intense peak at 577 °C, originating from the amorphous sericin extraction 59 and the degummed β-sheet fibroin degradation 59 .
The final ZrMOF@SF d hybrid was characterized by FT-IR and thermogravimetric analysis (TG-DTA) (see Fig. S1 in Supporting Information). The typical IR-peaks of SF are ≈1620, 1511, and 1226 cm −1 , characteristic for amide I (C=O stretching), amide II (N-H deformation, and C-N stretching), and amide III (C-N stretching and N-H deformation). FT-IR (cm −1 , selected peaks) ZrMOF@SF d : ν(OH); ν(C-H); ν(C=C); ν(C-OH); ν(C-Ο) (see Fig. S1 in Supporting Information). The TG-DTA curves for ZrMOF@SF d show a thermal-response profile similar to that of the SF d fibers. The TG% curve of ZrMOF@SF d exhibits a total weight loss of 97.3% in the range of 20-700 °C with a broad exothermic peak at 400-600 °C assigned to the fibroins' degradation. The DTA curve of ZrMOF@SF d presents a shifting of the degradation temperature around T = 360 °C (compared to the SF d DTA curve at T = 320 °C), which is attributed to the combustion of the organic part of ZrMOF@SF d (estimated w:w ≈ 5.7%).
The FT-IR spectra for ZrMOF and As III -loaded ZrMOF (Fig. 2d) are very similar, indicating the retention of the structure of the ZrMOF after As III sorption. Noteworthy, in the IR spectrum of As III @ ZrMOF, there is a band around 740 cm −1 and 1040 cm −1 , which is attributed to As III -O stretch from H 3 AsO 3 25,48,60,61 . X-ray photoelectron (XPS) analysis was used to determine the As III -valence state and the eventual interaction between arsenic and the adsorbent. The high-resolution As3d XPS spectrum, shown in Fig. 2e, clearly indicates that As III is the only oxidation form adsorbed onto ZrMOF sorbent. The characteristic peak at 44.2 eV corresponds to As III in agreement with Sudhakar et al. 62 , while no peak corresponding to As V is detected in As-loaded www.nature.com/scientificreports www.nature.com/scientificreports/ ZrMOF. This result shows that after adsorption of the As III on the ZrMOF, there is no oxidation event of As III , thusall the bound As atoms on ZrMOF are in the As III oxidation form. This information is in agreement with our FT-IR data, which detects the As III -O stretch, Fig. 2d, originating from H 3 AsO 3 . Also, the prevalence of the H 3 AsO 3 form is corroborated hereafter by the adsorption-isotherms' analysis, which shows that the adsorbed species is exclusively the neutral form of As III, i.e., H 3 As III O 3 .
ZrMOF is a highly porous material with a 12-connected net based on [Zr 6 O 4 (OH) 4 ] hexanuclear units interconnected via NH 2 -BDC 2− ligands. We should note that pZrMOF is charged due to the protonation of amine groups (as the ZrMOF is prepared in acidified water). Prior to the As III sorption investigations, the ZrMOF was treated with MeOH/Et 3 N to deprotonate the amine groups, thus resulting in a neutral framework.
A severe decrease of the Specific Surface Area of the ZrMOF is observed upon As III adsorption, see Fig. 3a,b. The non-linear [SSA vs. As III ] trend in Fig. 3b, for ZrMOF can be analysed into two different domains: [i] at low As III -loading the SSA is decreased moderately, [ii] at high-As III loading there is an abrupt lowering of the SSA. This change in SSA is not due to alteration of the crystal structure of ZrMOF, as verified by PXRD, Fig. 2b. Thus, the severe decrease of SSA upon As-uptake provides important insight into the As III -uptake mechanism by ZrMOF as follows: the SSA of 610 m 2 gr −1 for ZrMOF is equivalent to 6.1×10 20 nm 2 per gram of ZrMOF. The molecular volume of H 3 AsO 3 in H 2 O has been estimated by Canaval et al. 63 to be 75 ± 10Å 3 . Accordingly, each nm 2 surface element of ZrMOF can accommodate not more than 2 H 3 AsO 3 molecules, which gives an N-maximum of surface-adsorbed H 3 AsO 3 molecules N max = 2 [H 3 AsO 3 per nm 2 ] × [6.1×10 20 nm 2 per gram of ZrMOF] ~ 1.5mmoles of H 3 AsO 3 per gram of ZrMOF. When we compare this vs. the maximum As III -uptake capacity i.e. ~30 mmoles H 3 AsO 3 per gram, we conclude that the experimental As-uptake is 20 times higher than the maximum As III -uptake capacity of 1.5 mmoles of H 3 AsO 3 per gram, that would correspond to a mere surface coverage. Instead, the SSA drop vs. As-uptake data in Fig. 3b indicates a pore-filling mechanism, not a simple surface complexation. At the same time, the crystallinity of the ZrMOF material is retained after As III -uptake, see XRD in Fig. 2b. This result makes the ZrMOF behaving like an "As III -sponge" being capable of adsorbing unprecedented high-amounts, i.e., 2.2 grams of toxic As III per gram of ZrMOF.
As iii -adsorption kinetics. Kinetic data of As III adsorption by ZrMOF, Fig. 4a, show fast kinetics with a non-linear time-profile. The kinetic data can be fitted by the Weber and Morris model, described by Eq. 2 64 in 0 5 The Weber and Morris model 64 is based on the key-assumption that diffusivity and mass-transfer phenomena are determining the adsorption process. In Eq. 2, the fittable parameters are the kinetic constant rate K in (g g −1 h 1/2 ), and C (g g −1 ) which is a constant depending on the type of the boundary layer 64 . Accordingly, the data in Fig. 4a can be fitted by considering two different sets of K in and C, listed in Table S1 of the Supporting Information. At early adsorption times, (red circles in Fig. 4a), a fast kinetic constant K in = 6 g g −1 h 1/2 is obtained, with C = 0.11 g g −1 while at prolonged adsorption times, the kinetic constant is much lower K in = 1.9 g g −1 h 1/2 , with C = 1.1 g g −1 . This analysis reveals a two kinetic-phase phenomenon for As III uptake by ZrMOF. Taking into account the analysis of SSA data, we consider that the initial fast phase, corresponding to low-As uptake, is responsible for low decrease of SSA. At prolonged interaction times, where the adsorbed As III is high, a slower kinetic phase is operating, which corresponds to the sharp drop of SSA, i.e. the pores of ZrMOF are filled up with H 3 AsO 3 .
The same kinetic two-phase profile is observed for the ZrMOF@ SF d hybrid, Fig. 4b, indicating that the grafted ZrMOF particles operate similarly, i.e. surface adsorption of As III at low concentrations (red symbols in Fig. 4b) followed by pore filling at high As III -concentrations (green symbols in Fig. 4b). The silk fiber plays a minor role in As-uptake i.e. see adsorption isotherm in Fig. 5.
As iii -adsorption isotherms. A non-linear isotherm characterizes the uptake of As III by ZrMOF, see  www.nature.com/scientificreports www.nature.com/scientificreports/ As III -concentrations (<25 mg As III Lt −1 ), the isotherm shows a Langmuir-like trend, see solid-red symbols in Fig. 5( ). [ii] At increased initial As III concentrations, the isotherm data show a linear As-uptake isotherm, Fig. 5( ). This trend continues up to 75-80 mg of added As III Lt −1 . At even higher initial As III , the isotherm flattens, indicating a saturation of the As III -uptake by ZrMOF. Based on the data of Fig. 5, the maximum adsorbed As III per gram of ZrMOF material corresponds to a maximum of 2200 mg As III gr −1 of ZrMOF. The ZrMOF@SF d material exhibited a similar two-isotherms profile, see Fig. 5( , ) When normalized [per mass of grafted ZrMOF], the As III -uptake data in Fig. 5( , ) show that the performance of the ZrMOF@SF d material is within ~10% comparable to ZrMOF. Thus, grafting of ZrMOF retains its As III -uptaking capacity. For reference, pZrMOF and the SF d alone Fig. 5( ), (▲) show a very low As III -uptake, i.e. 0.260 g g −1 , and 0.068 g g −1 respectevelly. The significant inhibitory effect of the cation sites in cationic pZrMOF, to As III uptake i.e. vs the neutral ZrMOF, reveal that the surficial NH 2 sites play key role in the uptake mechanism. This is further analyzed in the following in the theoretical surface Complexation Modeling hereafter.
Theoretical modeling of the data in Fig. 5 can be done using two isotherm-adsorption models.
[a] At low added-As III concentrations, a Langmuir equation (Eq. 3) describes adequately the process, see fit (red line in Fig. 5)  Table S1. In (a) the added As III concentration was 50 mg L −1 . In (b) the added As III concentration was 15 mg L −1 . www.nature.com/scientificreports www.nature.com/scientificreports/ where q m (mg g −1 ) is maximum As III adsorption, q L ADS (mg g −1 ) is the surface concentration of adsorbed As III species in materials. C e (mg L −1 ) is the initial As-concentration. K Langmuir is the Langmuir stability constant representing the strength of As III -binding of the ZrMOF surface 65  where q Part gives the bound As III -moieties in the ZrMOF, in mg g −1 . The index n reflects a constant related to the intensity of sorption or the degree of the dependence of sorption on concentration. The efficiency of uptake is classified according to the value of K part . The linear Freundlich-type adsorption isotherms can be used to describe pore filling/partitioning processes 66 in hydrophobic/hydrophilic interfaces. Using the two isotherms Eqs. 3 and 4, the data in Fig. 5 can be fitted (see solid lines in Fig. 5) with the parameters listed in Table 1. More specifically, at low As III -concentrations, the uptake capacity -due to surface complexion-can achieve a maximum of q m = 0.83 g g −1 of As III at pH = 7. At high As III -concentration, where pore filling is operating, Fig. 5 (green line) a maximum As-uptake is attained of q e = 2.2 g g −1 of As III at pH = 7.
Further analysis of the surface adsorption for the As III species can be done by modeling of the pH-dependent As-uptake on the ZrMOF. This analysis, detailed in our previous works 16,24,67 , is described in Supporting  Information, Fig. S4. According to Fig. S4, the pH-dependent profile for low As III -concentrations shows that As III binds in its neutral form H 3 AsO 3 at the neutral amino-sites of neutral ZrMOF. This result is in agreement with Georgiou et al. 16 , Gupta et al. 27 , Su and Puls 28 . The surface amines ≡NH 2 act a specific binding sites for As III -uptake, see reaction (15), and (17) ≡NH 2 + H 3 AsO 3 ↔ ≡NH 2 -[H 3 AsO 3 ] in Table S2. We underline that in the cationic pZrMOF, the protonated ≡NH 3 + sites, reaction (12) in Table S2, do not favor adsorption of As III and this is the origin of the inferior performance of pZrMOF vs. ZrMOF. This is structurally described in Fig. S4d.
Since, in natural waters, several ions may coexist with arsenic, these can potentially compete with As-uptake 68 . In this context, the impact of competing ions such as PO 4 3− , CO 2− 3, NO 3 − , SO2− 4 , Cl − and HCO 3 − on As III adsorption was studied, see Fig. S5 in the Supporting Information. The data in Fig. S5, indicate that ZrMOF and ZrMOF@ SF d effectively remove As III even in the presence of CO 3, NO 3 − , , SO 4 2-, Cl − and HCO 3 − . The stronger inhibitory effect is exerted by PO 4 3− ions which may inhibit As III adsorption by 40% and 60% for ZrMOF and ZrMOF@SF d, respectively. The results are well agreement with Sudhakar 62 et al. and Jain and Loeppert 69 , which point out that the natural water ions do not affect the As III adsorption except for PO 4 3 which destabilizes the MOF structure.
Finally, we have evaluated the possibility of reusing the ZrMOF and ZrMOF@SF d materials after regeneration. Thus, we have applied the regeneration protocol 24,70 , which involved high ionic-strength treatment. More particularly, the protocol involves incubation for 24 hours under stirring at a pure aqueous solution containing 1 M of KNO 3 . Our data show that the As III adsorbed on either ZrMOF or ZrMOF@SF d cannot be removed by this treatment, indicating the high stability of the bound As III , i.e. attributed to its irreversible penetration into the pores of the ZrMOF.
Comparison of As iii -uptake with similar metal-organic framework materials. Figure 6 summarizes a comparison of As III sorption by the present materials vs. other pertinent MOF-based materials reported in the literature. According to Fig. 6, ZrMOF supersedes by far any known material.
This result is attributed to the fundamentally different mode of action of the neutral ZrMOF, i.e. the partitioning-like mechanism resulting in pore-filling allows exploitation of the full pore volume as a "sponge" for the uptake of the As III species form solution.

conclusions
Using XPS, FTIR, BET-porosimetry data, with theoretical Surface-Complexation-Modeling (SCM), we report a two-step phenomen non which boosts high-As ΙΙΙ -uptake. First, at low As III -concentrations, surface-complexation of H 3 AsO 3 results in As III -coated voids of neutral ZrMOF, and subsequently, at high As III -concentrations, the As III -coated voids of neutral ZrMOF are filled-up by H 3 AsO 3 via a partitioning-like mechanism. Also, we present an innovative hybrid-material, ZrMOF@SF d operating like an "As III -sponge" with unprecedented efficiency of 1800 mg As III  www.nature.com/scientificreports www.nature.com/scientificreports/ on SF d . ZrMOF itself exhibits As III adsorption of 2200 mg gr −1 , which supersedes any -so far-known As ΙΙΙ -sorbent. The reference materials i.e. cationic-pZrMOF and SF d play secondary role in As III -adsorption with adsorption capacity 260 mg As III gr −1 and 68 mg As III gr −1 respectively. Finally, the present research exemplifies for the first time a novel concept of a "partitioning-like" mechanism, operating for adsorption of H 3 AsO 3 , by neutral metal oxide materials. So far, such a mechanism has been conceptualized only for the uptake of non-polar organics by natural organic matter or synthetic polymers.

Methods
Materials. Sodium meta-arsenite NaAsO 2 was obtained from Sigma-Aldrich, while HCl, NaOH, The Silk Fabric (SF) provided by Tsiakiris Georgios Silk Company, Alexandroupoli, Greece. Sodium carbonate (Na 2 CO 3 ) purchased from Riedel de Haën. The coupling agent 3-(chloropropyl)trimethoxysilane was provided by Fluka. Methanol and ethanol purchased from Merck and diethyl ether from Sigma Aldrich.
All reagents were of analytical reagent grade purity, and all solutions prepared using deionized water obtained with a Milli-Q system with a conductivity of 18.2 μS cm −1 .
ZrMof preparation. The protonated (cationic) [Zr 6 O 4 (OH) 4 (NH 3 + -BDC) 6 ]Cl 6 • 35H 2 O (herein codenamed as pZrMOF) was synthesized as described previously 71 . To prepare the neutral material, pZrMOF (100 mg, 0.038 mmol) was treated with Et 3 N (72.  www.nature.com/scientificreports www.nature.com/scientificreports/ Physical characterization of materials. EDS analysis for ZrMOF showed no Cl confirming the complete deprotonation of ammonium groups. The powder X-ray diffraction (PXRD) measured at room temperature on an STOE-STADIMP powder diffractometer. PXRD equipped with an asymmetrically curved Germanium monochromator (CuKα1 radiation, λ = 1.54056 Å) and a one-dimensional silicon strip detector (MYTHEN2 1 K from DECTRIS). The line focused Cu X-ray tube operated at 40 kV and 40 mA. Powder of each sample was packed in a 1 mm diameter polyimide capillary (polymer substrate with neither Bragg reflections nor broad peaks above 10 °) and measured in Debye-Scherrer geometry on a spinning stage (~200 rpm). Intensity data from 3 to 125 degrees 2Θ collected for 17 h with a step of 0.005 degrees. The instrument calibrated against a NIST Silicon standard (640d) before the measurement.FT-IR spectra were recorded on KBr pellets in the 4000-400 cm −1 range using a Perkin-Elmer Spectrum GX spectrometer. Thermogravimetric analyses (TGA) were performed on a NETZSCH STA 449 C system. Sample analysis was conducted from 25 to 900 °C in an air atmosphere (50 mL min −1 flow rate) with a heating rate of 10 °C min −1 . Scanning electron microscope (SEM) performed by a JEOL JSM-6390LV equipped with an Oxford INCA PentaFET-x3 energy-dispersive X-ray spectroscopy (EDS) detector. Data acquisition performed with an accelerating voltage of 20 kV and 120 s accumulation time. Then the images were taken with a field emission JEOL JSM 7000 F electron microscope operating at 15 kV accelerated voltage. The samples were sputter-coated with 5-10 nm also reduce charging by used Au film.
X-Ray Photoelectron Spectroscopy (XPS) measurements were using a SPECS GmbH. Instrument equipped with a monochromatic MgKα source (hν = 1253.6 eV) and a Phoibos-100 hemispherical analyzer. The spectra were recorded under ultra-high vacuum s with a base pressure of 2-5 × 10 -10 mbar. Prior to measurement, the samples were placed on silicon substrates under high vacuum, before being placed in the main chamber for XPS measurement. The take-off angle was 45°. The recorded spectra were the average of three scans, with energy step 0.1-0.2 eV and a dwell time of 1 s. The As3d binding energy is calibrated based on the C1s core level at 284.5 eV. The spectral analysis included [i] a Shirley background subtraction, and [ii] peak deconvolution employing mixed Gaussian-Lorentzian functions in a least-squares curve-fitting program (WinSpec, Laboratoire Interdisciplinaire de Spectroscopie Electronique, University of Namur, Belgium) 72,73 . N 2 adsorption isotherms were measured at 77 K using a Quantachrome NOVAtouch LX 2 . Before analysis, all samples were degassed at 80 °C under vacuum (<10 -5 Torr) for 16 h. The specific surface areas were calculated by the Brumauer-Emmett-Teller (BET) method using the N 2 -adsorption data points, in the relative-pressure range P/P o of 0.05-0.35.
As iii analytical determination. The concentration of As III in the aqueous solution determined by square wave Cathodic Stripping Voltammetry (SW-CSV) using a Trace Master5-MD150 polarograph by Radiometer Analytica. SW-CSV is well suited for the analytical determination of As III 22,74 with a low detection limit (0.5 μg L −1 ). The measuring borosilicate glass cells obtained from Radiometer Analytica. The working electrode was a hanging mercury drop electrode (HMDE) with drop diameter of 0.4 mm generated by a 70 μm capillary. An Ag/AgCl electrode with a double liquid junction used as the reference electrode with a Pt measuring electrode. Importantly, samples were not purged with N 2 gas to avoid the loss of As III . During the stripping step, the solution stirred at 525 rpm. All measurements we performed using aliquots of 8.3 mL shifting at pH < 0.5 by 1.5 mL from 6.66 M of HCl and final 2 M concentration in the electrochemical cell, then 8 ppm of Cu 2+ were added. In the following, As III was determined by SW-CSV with accumulation potential E = −400 mV and accumulation time in the 60 s. As III quantified by its signal at E 1/2 = −670 mV 16,19,22 . As iii adsorption experiments. For the kinetic measurements, As III uptake from aqueous solutions studied in batch experiments. The kinetics of As III adsorption using ZrMOF studied as follows: 4 10 -4 g L −1 of ZrMOF were dispersed in 50 mL buffered aqueous solution in polypropylene tubes at pH 7, in the presence of 50 mg L −1 As III . For samples, ZrMOF @ SF d 3.4 mg was dispersed in 25 mL buffered aqueous solution in polypropylene tubes at pH 7, in the presence of 15 mg L −1 As III . The time-evolution of As III concentration was monitored at contact times ranging between 0-240 min and 0-960 min, respectively. At the end of each contract period, all sample centrifugation and the supernatant solution analyzed for As III . To determine the adsorption rates of As III , the amount of As III adsorbed at time t, q (mg As III g −1 ), calculated from the mass-balance between the initial concentration and the concentration at time t onto the solid adsorbents 16,29 .
Adsorption isotherms for pZrMOF and ZrMOF were recorded at pH 7 in the presence of 0-100 mg L −1 NaAsO 2 interacting with 0.1 g L −1 pZrMOF and 0 to 150 mg L −1 NaAsO 2 , interacting with 4 10 -4 g L −1 of ZrMOF suspended in 50 mL buffer solution in polypropylene tubes. On the other hand, for SFd and ZrMOF @ SF d , 0 to 100 mg L −1 and 0 to 150 mg L −1 NaAsO 2 and 0.2 g L −1 ,0.14 g L −1 respectively were suspended in 25 mL buffer solution in polypropylene tubes.
pH-dependent (pH-edge) experiments allow a detailed probing of the interfacial adsorption mechanisms 16,29 , while adsorption isotherms report the maximum uptake capacity. In this work, pH-edge adsorption experiments carried out for an initial concentration of 5, 5, and 15 mg L −1 (NaAsO 2 ) and also 0.1 g L −1 ,0.4 10 -4 g L −1 , 0.14 g L −1 of pZrMOF, ZrMOF and ZrMOF @ SF d respectively, suspended in 50 mL buffer solution whose pH adjusted in the range 4 to 8, in polypropylene tubes.
After metal addition, the suspension was allowed to equilibrate for 15 min ZrMOF, and pZrMOF at RT, while agitated using a magnetic stirrer. After completion of equilibration, the suspensions centrifuged at 6000 rpm for 10 min, and the supernatant solutions were analyzed for As III as described above. For SF d and ZrMOF @ SF d after metal addition, the suspension was allowed to equilibrate overnight at RT while using a magnetic stirrer. After completion of equilibration, the SF d or ZrMOF @ SF d suspension collects by metal tweezer.
Reuse experiments were also conducted for ZrMOF, which had adsorbed As III at pH 7. To reuse the samples, we had to desorb the adsorbed As. Thus, following the method used in ref. 30,75 , after As III adsorption, the material was immersed in an aqueous solution of 1 M KNO 3 for 24 h, and the supernatant was analyzed for As III