Guar Gum-Grafted Terpolymer Hydrogels for Ligand-Selective Individual and Synergistic Adsorption: Effect of Comonomer Composition.

Grafting of guar gum (GG) and in situ strategic attachment of acrylamidosodiumpropanoate (ASP) via solution polymerization of acrylamide (AM) and sodium acrylate (SA) resulted in the synthesis of a sustainable GG-g-(AM-co-SA-co-ASP)/GGAMSAASP interpenetrating polymer network (IPN)-based smart superadsorbent with excellent physicochemical properties and reusability, through systematic optimization by response surface methodology (RSM) for removal of methyl violet (MV) and/or Hg(II). The relative effects of SA/AM ratios, in situ allocation of ASP, grafting of GG into the AMSAASP terpolymer, ligand-selective superadsorption mechanism, and relative microstructural changes in individually/synergistically-adsorbed MV-/Hg(II)-/Hg(II)-MV-GGAMSAASPs were determined by extensive analyses using Fourier transform infrared (FTIR), proton nuclear magnetic resonance, ultraviolet-visible (UV-vis), and O 1s-/N 1s-/C 1s-/Hg 4f7/2,5/2-X-ray photoelectron spectroscopies, thermogravimetric analysis, differential scanning calorimetry, X-ray diffraction, field emission scanning electron microscopy, and energy-dispersive spectroscopy and were supported by % gel content, pHPZC, and % graft ratio. The ionic/covalent-bonding, monodentate, bidentate bridging, and bidentate chelating coordination between GGAMSAASPs and Hg(II), and MV+-Hg(II) bonding were rationalized by FTIR, UV-vis, fitment of kinetics data to the pseudo-second-order model, and thermodynamic parameters. The maximum adsorption capacities of 49.12 and 53.28 mg g-1 were determined for Hg(II) and MV, respectively, under optimized conditions.

Introduction

Hydrogels are porous polymeric networks produced by the cross-linking of synthetic and/or natural polymers (NPs) and can become many times voluminous by absorbing water during reversible swelling.[1] The relative availability of various hydrophilic/hydrophobic functional groups, originated via variation in synthesis parameters, such as temperature, initial pH (pHi) of solution, mole ratios of monomers, wt % of NP, cross-linker, and initiator, significantly affects the swelling properties of interpenetrating polymer network (IPN) hydrogels. In fact, grafting of NPs into the network of a homo/co/terpolymer, followed by cross-linking to the either polymer(s) to produce semi/full-IPN, can provide the optimum balance between mechanical strength and biodegradability.[2,3] Synthetic/semisynthetic hydrogels are mostly used in tissues, chromatography, pharmaceutical/cosmetic industries,[4] and sustained drug delivery. Hydrogels are used in artificial cornea implants[5] and as biosensors for hepatitis B antigen,[6] glucose,[7] hemoglobin A1c (HbA1c),[8] and bile acids.[9]

Nowadays, hazardous dyes/metal ions [M(II/III/VI)] of waste effluents cause serious threat to all living organisms via severe water pollution.[10] The most important anthropogenic sources of Hg evolution in aquatic systems involve atmospheric deposition, erosion, urban discharges, agricultural materials, mining, combustion, industrial discharges, and compact fluorescent lamps.[11−13] Mercury is utilized in chlor-alkali and vinyl chloride monomer production, artisanal small scale gold mining, gold recovery from electronic waste, pesticides, and many other industrial applications. The United States Environmental Protection Agency has suggested 0.002 mg L–1 and 10 μg L–1 discharge to be the maximum allowable limit for Hg(II) in drinking and waste water, respectively. Therefore, it is imperative to eliminate even trace amounts of such a serious pollutant before discharging effluents into the environment, for human safety and environmental protection. Hg inflicts severe toxicity due to the nonbiodegradability and recirculating property in the environment.[14,15] In addition, Hg causes neurologic malfunction, gastrointestinal disorder, and renal problems.[14] Methylmercury, the most toxic form of Hg, is highly detrimental for human embryo even when present in very small amounts.[16]

Meanwhile, more than 10 000 different types of synthetic aromatic dyes are used as coloring agents in plastic, pharmaceutical, textile, printing, paper, food, and other chemical industries.[17] Industrial wastes containing of about 15 wt % nonbiodegradable dyes and pigments eventually impart severe carcinogenic effects. The modern techniques used for eliminating industrial waste effluents are reverse osmosis, reduction, coagulation,[18] solvent extraction, ion exchange,[18] chemical precipitation, membrane-based separation,[19−21] and adsorption.[22,23] Of these, adsorption is preferred because of the versatility, cost effectivity, simplicity of design, accuracy, selectivity, and high degree of efficiency.[13] Indeed, the suitability of the triphenylmethane dyes for M(II) detection[24] has eventually encouraged the simultaneous removal of Hg(II) and methyl violet (MV).

Guar gum (GG), a nonionic hydrophilic galactomannan polysaccharide seed gum, contains linear chains of a (1–4)-linked β-d-mannopyranose backbone with branch points from their 6-positions linked to β-d-galactose (i.e., 1–6-linked β-d-galactopyranose) in a 1:2 ratio.[25] This O–H rich galactomannan forms H bonds, imparting significant viscosity. Moreover, GG finds high-performance applications in a large number of industries because of solubility in cold water, gelling properties, pH stability from 4.00 to 10.50, film-forming ability, biodegradability, nontoxicity, low cost, and renewability. However, GG-based IPNs not only bring the favorable properties by introducing newer functional groups but also impart the inherent diversified advantages of GG. Indeed, chemical modification of GG removes their intrinsic deficiencies, which may restrict their overall utilization in diversified field of applications. In this context, the graft copolymers of GG, such as GG-g-poly(ethylacrylate),[26] GG-g-poly(acrylic acid) (AA),[27] GG-g-poly(acrylamide) (AM),[28] and GG-multiwalled carbon nanotubes nanocomposite hydrogels,[29] have effectively been used for adsorptive removals of various aqueous waste contaminants. Again, graft copolymers, such as potato starch-g-poly(AM-co-AA-co-bicarboxylic itaconic acid),[30] xanthan gum-g-poly(hydroxyethylmethacrylate-co-AA),[31] quaternary ammonium chitosan-g-poly(acrylic acid-co-acrylamide),[32] and GG-g-poly(AM-co-aniline),[33] have been tested for removal of dyes and M(II/III/VI) such as Cu(II), Ni(II), Pb(II), Zn(II), Fe(II), Cr(VI), and Hg(II). In this context, incorporation of an acrylamido functionality in the polymer matrix was carried out using several acrylamide derivatives, such as 2-acrylamido glycolic acid,[34] 2-acrylamido-2-methylpropane sulfonic acid,[35] 2-acrylamido-2-methyl-1-propanesulfonic acid,[36] and 3-acrylamido phenylboronic acid,[37] as external monomers. In this context, we carried out in situ incorporation of acrylamidosodiumpropanoate (ASP) within the network of GG-grafted AMSAASP (i.e., GGAMSAASP) via solution polymerization of sodium acrylate (SA) and AM, which is yet to be reported. Meanwhile, several homo/copolymers, IPNs, and composite hydrogels have been reported for removal of dyes and M(II) individually.

The synthesis of tailor-made sustainable GGAMSAASP, an NP-grafted terpolymer, without ex situ incorporation of ASP, the introduction of new possible pathway for superadsorbent synthesis, adjunct allocation of ASP, extensive microstructural analyses of GGAMSAASP14/18, MV–GGAMSAASP14/18, Hg(II)–GGAMSAASP14/18, and Hg(II)–MV–GGAMSAASP14/18, effect of variation in the composition of GGAMSAASPs and ligand-selective MV, and/or Hg(II) superadsorption mechanism influencing relative distribution of Hg(II) at the surface/bulk have extensively been studied and reported for the first time.

Experimental Section

Materials

AA, AM, GG, N,N′-methylenebisacrylamide (MBA), potassium persulfate (PPS), sodium bisulfite (SBS), MV, and HgCl2 of analytical grades were purchased from Sigma-Aldrich and used without any further modification.

Synthesis of Hydrogels

GGAMSAASP was synthesized through free radical solution polymerization via grafting of GG to the terpolymer network and in situ adjunct allocation of ASP, SA, and AM, employing PPS and SBS as the redox initiator system and MBA as the cross-linker in a N2 atmosphere. The exact composition of ingredients and the reaction temperature were preoptimized by response surface methodology (RSM) via ensuring maximum swelling of the synthesized hydrogels, which were prepared by successive incorporation of varied dosages of AM and MBA at different pHi values. Finally, two different GGAMSAASPs, one possessing the optimum conditions (GGAMSAASP18, AM/SA = 1:8) and the other containing a different AM/SA ratio (GGAMSAASP14, AM/SA = 1:4), were chosen for the ligand-selective adsorption study. The motive for synthesizing different GGAMSAASPs of varying compositions was to investigate the relative distribution and population of Hg(II) and MV, both on the surface and in the bulk of the adsorbents. Initially, two uniform suspensions, each containing 0.5 g of GG powder, were prepared in 42.9/46.2 mL of H2O at 313 K using an ultrasonicator to ensure complete dissolution of GG without formation of clumps. Then, in a three-neck reactor, the as-obtained GG suspension was taken, followed by the dropwise addition of SA (0.26/0.21 mol prepared in 34.60/27.30 mL of water at the rate of 30 drops min–1) at pHi = 5.5 and addition of AM (0.02/0.07 mol, prepared in 8.00/12.00 mL of water) with constant stirring at 300 rpm. Then, the MBA solution (0.26 mmol prepared in 4.5 mL of water) was added at a constant temperature (i.e., 295 K), and the resultant solution was allowed to homogenize under a N2 atmosphere for 8 h. Thereafter, polymerization was initiated via gradual addition of PPS and SBS as redox initiators, prepared in 10 mL of water by dissolving 0.23 and 0.60 mmol, respectively. The prepared GGAMSAASPs were allowed to swell in 1:3 methanol/water solution (v/v) and washed several times for the complete removal of unreacted monomers and water-soluble oligomers. Finally, GGAMSAASPs were air-dried for 3 days, followed by drying under a vacuum oven at 323 K.

Characterization

Unadsorbed and/or adsorbed GGAMSAASPs were characterized by Fourier transform infrared (FTIR) through a Spectrum-2 spectrometer, Singapore, proton nuclear magnetic resonance (1H NMR) using a Bruker AVANCE Digital 300 MHz spectrometer in CDCl3 solvent with tetramethylsilane as an internal reference, X-ray photoelectron spectroscopy (XPS) via ESCA+, Omicron NanoTechnology, Oxford Instruments Germany, thermogravimetric analysis (TGA) using a Pyris6 TGA analyzer, The Netherlands within 30–700 °C, differential scanning calorimetry (DSC) via a Pyris6 DSC calorimeter, The Netherlands within 30–442 °C, X-ray diffraction (XRD) by an X’Pert PRO, PANalytical B.V., The Netherlands, field emission scanning electron microscopy (FESEM), and energy-dispersive spectroscopy (EDX) using a ZEISS EVO-MA 10 spectrometer. GGAMSAASP was also characterized by the % graft ratio, % gel content, pH at point of zero charge (i.e., pHPZC), and equilibrium swelling ratio (ESR) at different pHi values. RSM and all graphics based analyses were carried out using Design-Expert 7.0.0 and Origin 9.0 software.

Experimental Design for RSM-Optimized Synthesis of GGAMSAASP18

Hydrogels exhibit variable swelling by the prevalence of diversified hydrophilic functional groups. In fact, the relative population of such groups predominantly depends on the change in the amounts of synthetic parameters, such as AM (wt %, A), initiator (wt %, B), GG (wt %, C), cross-linker (wt %, D), and pHi (−, E). Because the hydrogel possessing the maximum ESR is anticipated to impart the highest adsorption capacity (AC) toward removals of dyes/M(II), identification of the optimum conditions for synthesizing such a hydrogel is essential. In fact, synergistic effects along with individual effects of such process variables may play a pivotal role in the variation of ESR of hydrogels, and hence, optimization of the process variables is crucial. However, investigation of such individual and/or synergistic effects through one factor at a time (OFAT) approach is very tedious and does not include the two factor interaction(s), that is, 2FI(s), between variables. Therefore, the design of experiments method was used, which included individual and/or synergistic effects and reduced the overall cost of research via substantial reduction in the total number of experiments. In this context, RSM, a frequently used statistical method, was utilized because it involves both OFAT and 2FI approaches. However, the use of a full factorial RSM design is highly strenuous because this consists of total 25 = 32 experimental runs. Therefore, fractional factorial design (resolution-IV), an intelligent RSM design, was adopted to isolate distinct variables incorporating the maximum effect on ESR, via analyzing Pareto chart and half-normality plots, for significant reduction in the experimental runs. Finally, the full factorial design was carried out by using the as-obtained most significant process variables through the central composite design (CCD) analysis.

Results and Discussion

FTIR

Several characteristic peaks of GG and AMSAASPs were noted to disappear in the FTIR spectra of GGAMSAASPs, together with notable alterations in several symbolic peaks of AMSAASPs (Figure a, S1), which indicated the intimate interactions between GG and AMSAASPs. For instance, the symbolic GG peak at 870 cm–1, designated to C–C/C–O vibrations coupled with anomeric C–H of β-conformers, disappeared in both the GGAMSAASPs.[38] Interestingly, besides the orthodox (1–4)/(1–6) glycosidic linkages, the formation of −CH2–O–CH2– type ether linkage within GGAMSAASPs was realized from the arrival of a new peak at 1122 cm–1, which was neither present in GG nor in AMSAASPs, confirming the grafting of GG within AMSAASPs via interactions between unsaturated SA/AM moieties and −CH2–O• generated from the primary alcohol of GG (Scheme ).[39] In fact, such conversion of −CH2OH into −CH2–O• was also substantiated by the complete loss of the symbolic −CH2OH peak of GG at 1076 cm–1 in both the GGAMSAASPs, together with appearance of a new C–O–C def. peak of aliphatic ether at 424/426 cm–1.[40] In fact, grafting-driven new ether linkage formation led to considerable alterations in various −CH2– vibrations, including the disappearance of the −CH2– twisting peak at 1024 cm–1 in GGAMSAASPs. Furthermore, the intimate phase mixing within GGAMSAASPs resulted in significant changes in the H-bonding environment, as apprehended from the arrival of new H-bonded O–H str. peaks in the range of 3600–3650 cm–1 for GGAMSAASPs. Moreover, several changes in the H-bonding environment were also realized from the arrival of new peaks at 2352, 2346, and 2331 cm–1 in GGAMSAASP18,[2] along with the marginal shifting of the broad peak from 3433/3435 cm–1 of AMSAASP18/14 to 3434 cm–1 in both GGAMSAASPs (Figure a). Moreover, in-plane vibrations of both C–C=O def. and N–C=O in primary amides were influenced by the grafting-related alterations in the surrounding H-bonding environment. For instance, the C–C=O def. peak at 481 cm–1 of AMSAASPs disappeared completely in both the GGAMSAASPs. Moreover, significant alterations in the H-bonding environment around AMSAASP moieties within GGAMSAASPs and the associated intermingling removed several characteristic bending peaks at 670, 657, 645, and 632 cm–1 of N–C=O in secondary amides, together with the formation of new N–C=O peaks at 674/672 and 651/648 cm–1 for GGAMSAASP18/14 (Figure a). Furthermore, changes in the vibrations of both C–C=O and N–C=O moieties should closely be associated with the notable decrease in C–N str. and bending vibrations of the amide III band from 1323 cm–1 for AMSAASP to 1319 cm–1 in GGAMSAASP14, whereas the same band remained unchanged at 1318 cm–1 in GGAMSAASP18, owing to the relatively fewer population of C–N bonds in GGAMSAASP18.[41] Indeed, several symbolic peaks, corresponding to amide III, C–C=O, and N–C=O of amides, indicated the transformation of primary into secondary amides via addition reaction between −CO–NH• and unsaturated chains to produce −CO–NH–CH2– segment, which eventually altered the prevalent −CH2– and C–N peaks, especially in GGAMSAASP14. Notably, the simultaneous existence of both −COOH and −COO– in GGAMSAASP18 was detected from the prevalent peaks corresponding to C=O str. of −COOH at 1742, 1738, 1718, and 1705 cm–1 along with νs(−COO–) and νas(−COO–) at 1405 and 1560 cm–1, respectively.[42] Indeed, the peaks at 1705 and 1742 cm–1 were attributed to C=O str. of cyclic H-bonded −COOH in dimeric form and free −COOH of GGAMSAASP18, respectively.[43] Though both −COOH and −COO– were simultaneously present in GGAMSAASP18, relatively fewer number of C=O str. peaks within 1700–1750 cm–1 and the complete disappearance of the peak at 1705 cm–1 in GGAMSAASP14 could be related to the complete absence of the cyclic H-bonded −COOH dimer because of the lowering of available −COOH in GGAMSAASP14. Again, the participation of −COO– in ionic bonding was ascertained from Δν, between νas(−COO–) and νs(−COO–), of 155/157 cm–1 for GGAMSAASP18/14, which was well within the range of ionic interaction (i.e., 136–164 cm–1).[44]

Figure 1

FTIR of (a) GGAMSAASP18/14, AMSAASP18/14, and GG and (b) Hg(II)–MV–GGAMSAASP18/14, Hg(II)–GGAMSAASP18/14, and MV–GGAMSAASP18/14.

Scheme 1

Synthesis of GGAMSAASPs

Notably, Hg(II) demonstrated preferential attachment with the amide over −COO–. Indeed, the intimate association between Hg(II) and amides could be anticipated by the formation of the Hg–N covalent bond, related to the newly appeared characteristic peaks at 507/511 cm–1 in Hg(II)–GGAMSAASP18/14. In this context, the complete disappearance of peaks at 671 and 648 cm–1, designated to N–C=O in-plane bending of α-branched aliphatic secondary amides, could be related to the formation of the covalent bond between Hg(II) and amides. In this context, Hoang et al. established the formation of the N–Hg–N bond between thymine bases of DNA via releasing two imino protons into the solution.[45] As suggested by the earlier works,[46] possible entrapment of Hg(II) within GGAMSAASPs might produce numerous >N–Hg–N< cross-links to impart increased steric hindrance, especially at the outer region of GGAMSAASP18, resulting in the substantial restrictive penetration and diffusion of Hg(II) from the densely cross-linked outer region to the bulk of Hg(II)–GGAMSAASP18. Furthermore, a relatively higher proportion of −COO– in GGAMSAASP18 should impart more hydrophilic character in GGAMSAASP18 over GGAMSAASP14. At the time of Hg(II) adsorption from the aqueous phase, more hydrophilic −COO– should preferentially be located on the hydrogel surface, more so in GGAMSAASP18. Thus, a relatively increased number of −COO–, especially on the GGAMSAASP18 surface, could interact with the incoming Hg(II) and thus resist the penetration into the bulk of the network. Moreover, the preferential binding of Hg(II) on the GGAMSAASP18 surface increased the hydrophobic character of the outer surface and thus resulted in the prevalence of some free −NH2 in the interior region. In fact, the presence of unreacted −NH2 could be detected in Hg(II)–GGAMSAASP18 from the appearance of a significantly intense peak at 1614 cm–1 (Figure b, S1), ascribed mainly to amide II, as a result of a typical bending vibration of free −NH2 along with νas(−COO–). In this regard, the characteristic peak for −NH2 of pure poly(AM) appears at 1613 cm–1.[47] However, a broad peak of relatively lower intensity at 1617 cm–1 was observed in Hg(II)–GGAMSAASP14, which suggested relative scarcity of free −NH2 in Hg(II)–GGAMSAASP14 as compared to that in Hg(II)–GGAMSAASP18. Nevertheless, small peaks at 667/1570 and 666/1570 cm–1 were attributed to −NH2def. and sciss., respectively, for free −NH2 that appeared in Hg(II)–GGAMSAASPs (Figure b).[40,48] Moreover, both the amide I bands of primary and aliphatic secondary amides within 1650–1670 and 1650–1630 cm–1, respectively, were not found in Hg(II)–GGAMSAASP18. On the contrary, the presence of both of these amide peaks in Hg(II)–GGAMSAASP14 indicated the lesser interaction of >C=O of amides with Hg(II). As the population of free −NH2 was relatively higher in Hg(II)–GGAMSAASP14, such −NH2 might be attacked by Hg(II) to produce >N–Hg–N< cross-links and −Hg–NH– moieties more frequently in Hg(II)–GGAMSAASP14 than in Hg(II)–GGAMSAASP18. Thus, in addition to the significantly higher proportion of −OH, relatively greater availability of free −NH2 in Hg(II)–GGAMSAASP18 ensured the prevalence of higher number of mutual weaker H-bonds among −NH2 and −OH to produce a broad and relatively intense peak at 3522 cm–1. Such an increase in the mutual H-bonding restricted the probability of strong H-bonding among individual O–H, as envisaged from the small peaks of lower intensities at 2352 and 2343 cm–1 in Hg(II)–GGAMSAASP18 than that of the relatively intense peaks of GGAMSAASP18. Accordingly, a broad but comparatively less intense peak at an appreciably lower wave number of 3434 cm–1 indicated lesser possibility of mutual −NH2/O–H H-bonding within Hg(II)–GGAMSAASP14. In fact, the lowering in such mutual H-bonding could encourage the formation of stronger H-bonds among O–H in Hg(II)–GGAMSAASP14, which is also reflected from the arrival of small intense peaks at 2352 and 2341 cm–1 (Figure b). However, in the presence of adsorbed Hg(II), the appreciable change in mutual H-bonding in Hg(II)–GGAMSAASP18 could be in consequence to the replacement of Na+ by Hg(II) during adsorption, producing covalent bonds between −NH2/–NH– of primary/secondary amides and Hg(II), along with the formation of ionic (I) and monodentate (M)/bidentate bridging (BB- type coordinate bonds within the increasingly available −COO– of GGAMSAASP18 with adsorbed Hg(II). Indeed, −COO– of GGAMSAASP18 formed coordinate bonds with Hg(II), preferentially via the M mode, that could also be realized from the appreciably intense νas(−COO–) at 1614 cm–1 and characteristic Δν of 210 cm–1 in Hg(II)–GGAMSAASP18 (Table ).[44] However, lesser probability of the BB mode between Hg(II) and −COO– of Hg(II)–GGAMSAASP18 was also envisaged from the appearance of a shoulder at 1558 cm–1, attributed to νas(−COO–). On the contrary, interactions between Hg(II) and −COO– in Hg(II)–GGAMSAASP14 were actuated preferably via the BB mode instead of the M mode (Table ). Additionally, bidentate chelating (BC) mode was also observed in Hg(II)–GGAMSAASP14, suggesting the prevalence of diversified interactions among Hg(II) and −COO– in Hg(II)–GGAMSAASP14. It was believed that the possible penetration of Hg(II) into the core of GGAMSAASP14 led to a more uniform distribution of Hg(II) throughout the surface and the bulk of Hg(II)–GGAMSAASP14, and hence, variegated interactions between Hg(II) and −COO– were feasible in Hg(II)–GGAMSAASP14 (Scheme ). In fact, better penetration of Hg(II) into the network of GGAMSAASP14 was also rationalized from the occurrence of new peaks at 3617, 3606, 3593, and 3574 cm–1, suggesting the alteration in O–H str. because of hampering of mutual H-bonding of O–H and primary amides, as primary amides were involved in the covalent bonding with adsorbed Hg(II) in Hg(II)–GGAMSAASP14. On the contrary, relatively restricted penetration of Hg(II) in GGAMSAASP18 was envisaged through the arrival of fewer number of new peaks at 3634 and 3592 cm–1 in Hg(II)–GGAMSAASP18. Though the attachment tendency of Hg(II) with O-donors was less preferred, deposition of insoluble Hg(OH)2 at pHi > 5.0 on both the GGAMSAASP surfaces was determined via appearance of small peaks at 3827/3828 cm–1 in Hg(II)–GGAMSAASP18/14.[49]

Table 1

Possible Modes of Interactions of Hg(II) and MV with −COO– of GGAMSAASPs

sample	νas(−COO–) – νs(−COO–) = Δν (cm–1)	mode(s) of interaction
GGAMSAASP18	1560 – 1405 = 155	I
GGAMSAASP14	1560 – 1403 = 157	I
Hg(II)–GGAMSAASP18	1614 – (1404/1452) = 210/162	M, I
	1558 – (1404/1452) = 54/106	I, BB
Hg(II)–GGAMSAASP14	1617 – (1402/1451) = 215/166	M, BB
	1578 – (1402/1451) = 176/127	BB
	1561 – (1402/1451) = 159/110	I, BB
	1558 – (1402/1451) = 156/107	I, BB
	1550 – (1402/1451) = 148/99	I, BB
	1542 – (1402/1451) = 140/91	I, BC
MV–GGAMSAASP18	1557 – 1406 = 151	I
MV–GGAMSAASP14	1558 – 1404 = 154	I

Scheme 2

Relative Distribution of Hg(II) in (a) Hg(II)–GGAMSAASP14 and (b) Hg(II)–GGAMSAASP18

Individual MV adsorption onto GGAMSAASPs affected the H-bonding environment within GGAMSAASP18, as realized from the changes in the O–H str. vibrations involved in both strong and weak H-bonding. In this context, the peak at 3655 cm–1 of GGAMSAASP18 disappeared (Figure b), together with the appearance of a new peak at 3593 cm–1 in MV–GGAMSAASP18, in consequence to the deprotonation of numerous −COOH into −COO– at alkaline pHi. Adsorption of MV and the accompanied conversion of −COOH into −COO– resulted in an appreciable reduction in the overall intensities and positions of the peaks corresponding to strongly H-bonded O–H str. in MV–GGAMSAASP18. Especially, a couple of relatively intense peaks at 2352 and 2346 cm–1 of GGAMSAASP18 were shifted to produce peaks of appreciably lower intensities at 2356 and 2350 cm–1 in MV–GGAMSAASP18. On the contrary, relatively feeble changes were manifested in MV–GGAMSAASP14. However, O–H str. peaks at 3645, 3564, and 3542 cm–1 of GGAMSAASP14 disappeared during MV adsorption. In fact, the conversion of −COOH into −COO– in MV–GAMSAASP18 was relatively intense and thus considerably shifted the νas(−COO–) and νs(−COO–) peaks to 1558 and 1406 cm–1, respectively, than that in GGAMSAASP18. Moreover, only I-type interactions between MV cations (MV+) and −COO– in MV–GGAMSAASPs was confirmed from the Δν values of 151 and 154 cm–1 lying well within 136–164 cm–1. Followed by MV adsorption, almost similar but lesser prominent changes were observed in MV–GGAMSAASP14 because of lower I-interactions between MV+ and relatively fewer number of −COO– in MV–GGAMSAASP14. In this context, adsorption of MV+ on both the GGAMSAASPs was also apprehended from the arrival of distinct and intense peaks at 1321/1322 cm–1 in MV–GGAMSAASP18/14, corresponding to the C–N str. vibrations of the terminal saturated dimethylamino groups of MV.[50]

Several significant changes, especially in the H-bonding environment, were observed during the combined adsorption of Hg(II) and MV. Again, during Hg(II)–MV synergistic adsorption, reappearance of GG specific peaks at 3645 and 3564 cm–1 (Figure b), lost earlier during individual MV or Hg(II) adsorption, could be explained by the mutual bonding between MV and Hg(II) via conversion of some free functional groups of both adsorbates to the bound state, which resulted in some unoccupied functional groups of the adsorbent, as realized from retaining of those O–H str. peaks within Hg(II)–MV–GGAMSAASP14. This signified the prevalence of a strong Hg(II)–MV interaction that ultimately weakened the adsorbent–adsorbate interaction. Mutual Hg(II)–MV interaction in Hg(II)–MV–GGAMSAASP18 also obliterated several less intense O–H str. peaks within 3500–3700 cm–1, present earlier in Hg(II)–GGAMSAASP18 and MV–GGAMSAASP18. Moreover, complete disappearance of the Hg(OH)2 specific peak at 3827 and 3828 cm–1 in both the Hg(II)–MV–GGAMSAASPs suggested the role of Hg(II)–MV interaction in arresting the formation of Hg(OH)2. Furthermore, mutual Hg(II)–MV interaction in Hg(II)–MV–GGAMSAASPs resulted in significant changes in C–N str. of terminal saturated dimethylamino groups of MV, realized from the significant shifting of peaks from 1321/1322 cm–1 in MV–GGAMSAASP18/14 to 1315/1318 cm–1 in Hg(II)–MV–GGAMSAASP18/14 (Figure b). Indeed, such remarkable shifting of peaks toward a lower frequency might be possible in consequence to the formation of coordinate bonding between Hg(II) and −NHMe/–NMe2 of MV. Earlier, Jia et al. reported coordinate bonding between −NMe2 of malachite green and Hg(II) to produce a Hg(II)–thymine-rich DNA–MG complex, wherein Hg(II) bridged the two thymine bases.[46] In fact, mutual interactions between Hg(II) and MV carried out considerable shifting of the characteristic Hg–N peak from 507/511 cm–1 in Hg(II)–GGAMSAASP18/14 to 510/516 cm–1 in Hg(II)–MV–GGAMSAASP/14.

NMR Analyses

AM

5.70 (1H, dd), 6.19 (1H, m), 6.30 (1H, m), 6.56 (2H, s) (Figure S2a); 13C NMR (75 MHz, CDCl3, δ, ppm): 127.42, 130.25, 168.10 (Figure S2b).

AA

1H NMR (300 MHz, CDCl3, δ, ppm): 6.03 (1H, dd), 6.23 (1H, dd), 6.59 (1H, dd), 11.41 (1H, s) (Figure S2c); 13C NMR (75 MHz, CDCl3, δ, ppm): 127.87, 132.97, 171.62 (Figure S2d).

MBA

1H NMR (300 MHz, DMSO-d6, δ, ppm): 4.49 (2H, t), 5.59 (2H, dd), 6.13 (4H, m), 8.73 (2H, t) (Figure S2e); 13C NMR (75 MHz, DMSO-d6, δ, ppm): 42.76, 125.53, 130.88, 164.41 (Figure S2f).

GG

1H NMR (300 MHz, DMSO-d6, δ, ppm): 3.67, 3.83–3.87, 4.01–4.06, 4.24, 4.75, 5.05, 5.28, 5.40 (Figure S2g).

AMSAASP

1H NMR (300 MHz, CDCl3, δ, ppm): 0.81, 1.22, 1.55, 2.06, 2.18, 2.32, 2.60, 3.45, 3.61, 4.25, 6.85, 7.72, 8.04 (Figure a).

Figure 2

1H NMR of (a) AMSAASP and (b) GGAMSAASP.

GGAMSAASP

1H NMR (300 MHz, CDCl3, δ, ppm): 0.81, 1.21, 1.56, 2.06–2.80, 3.45, 3.61, 3.87, 4.66, 5.33, 5.40, 6.87, 7.56, 8.04 (Figure b).

The prevalence of methylene (−CH2−) and methine (>CH−) linkages in the backbone of AMSAASP was determined through the appearance of characteristic peaks (Figure a) within 0.81–1.55 and 2.06–2.60 ppm, respectively. However, the symbolic peaks of −CH2– and >CH– protons in GGAMSAASP14 appeared within 0.81–1.56 and 2.06–2.80 ppm, respectively, suggesting the existence of relatively variegated >CH– protons as a consequence of the diversified branching network in GGAMSAASP14 (Figure b).[51] In addition, complete disappearance of several characteristic peaks of vinyl protons within 6.03–6.59, 5.70–6.30, and 5.59–6.13 ppm for SA, AM, and MBA, (Figure S2), respectively, indicated the comprehensive attainment of copolymerization reaction between SA and AM along with the cross-linking via MBA. Indeed, such cross-linking was rationalized through the appearance of characteristic −CH2– peaks at 4.25 and 4.66 ppm in AMSAASP and GGAMSAASP14, respectively.[2] In addition, the arrival of new peaks from 3.45 to 3.61 ppm in the spectra of both AMSAASP14 and GGAMSAASP14 confirmed the presence of −CH2– of −CH2–NH–(C=O)–,[52] appearing through the free radical chain propagation reaction between −C(=O)NH• and SA. Furthermore, −NH– of cross-linked MBA, newly formed moieties, and amide appeared at 6.85, 7.72, and 8.04 ppm in AMSAASP and 6.87, 7.56, and 8.04 ppm in GGAMSAASP14,[2] respectively. However, GGAMSAASP14 was also associated with some distinct peaks at 3.87, 3.67, and 5.40/5.33 ppm for galactose-5/mannose-4 (gal-5/mann4), gal-/mann-6, and anomeric protons of GG, respectively.[53,54] In fact, the grafting of AMSAASP through −CH2OH of GG was also confirmed from the significant shifting of the characteristic −CH2–O peak from 3.67 of GG to 3.61 ppm in GGAMSAASP14.[53,54]

XPS Analyses

GGAMSAASP14 and Hg(II)–GGAMSAASP14 were thoroughly analyzed by XPS to determine the adsorption mechanism and variegated interactions between Hg(II) and GGAMSAASP14 from the relative changes in the binding energies (BEs) and peak intensities. In this context, the deconvoluted O 1s spectrum of GGAMSAASP14 showed three characteristic peaks with BEs of 530.82, 535.97, and 540.73 eV (Figure c) for O=C of −COOH,[55]O–H of −COOH,[56] and O-atom shake-up satellite band of −CO–NH–CH2– in GGAMSAASP14, respectively. Earlier, the presence of the −CO–NH–CH2– segment in GGAMSAASP14 was also ascertained from 1H NMR and FTIR analyses. However, the O 1s peak at 540.73 eV indicated the prevalence of low-dimensional H-bonded water clusters around GGAMSAASP14.[57] In this context, the O 1s peaks at 530.82 and 535.97 eV of GGAMSAASP14 were shifted to 531.80 and 536.02 eV (Figure f), respectively, indicating conversion of −COOH into −COO– and inferior interactions of Hg(II) with O–H of −COOH, respectively. In addition, the O 1s peak at 540.73 eV of the low-dimensional H-bonded water cluster was significantly shifted to 540.01 eV in Hg(II)–GGAMSAASP14 because of the change in H-bonding, also apprehended earlier from the FTIR analysis. Moreover, the deconvoluted C 1s peaks at 284.96 and 285.76 eV (Figure a) were ascribed to C–C[55] and C=O of amide/acid fragments[58] in GGAMSAASP14, respectively. As C-ends of C–C and C=O were indirectly involved in Hg(II) adsorption, these C 1s peaks were shifted slightly to 284.95 and 285.92 eV (Figure d), associated with the lowering in the respective peak intensities. Furthermore, the N 1s peak at 399.95 eV (Figure b) was attributed to the prevalent O=C–NH2 in GGAMSAASP14. In fact, a significant shifting of the N 1s peak form 399.95 to 399.35 eV (Figure e) was observed in Hg(II)–GGAMSAASP14 as a consequence of Hg(II) adsorption. In fact, the formation of a covalent bond between the N-end of O=C–NH2 and Hg(II), such as O=C–HN–Hg(II)–NH–C=O, was ascertained from the lowering of such a BE. Additionally, coordinate bonding between Hg(II) and N-donors was also rationalized from the significant increase in the N 1s BE up to 403.11 eV. As adsorption of Hg(II) was mostly governed by the greater relative enhancement of primary amides, higher ACs were also noted in GGAMSAASP14. Additionally, BEs of both Hg 4f7/2 and Hg 4f5/2 substantially decreased to 99.49/100.60 and 103.52/104.70 eV (Figure g), respectively, in comparison to the characteristic BEs at 102.58/106.68 eV of Hg 4f7/2/Hg 4f5/2 for HgCl2.[59] In fact, surface deposition of HgCl2 on GGAMSAASP14 was also apprehended from the appearance of Hg 4f7/2 BE at 102.20 eV.

Figure 3

XPS analyses of C 1s (a/d), N 1s (b/e), and O 1s (c/f) for GGAMSAASP14/Hg(II)–GGAMSAASP14; (g) Hg 4f7/2 and 5/2 for Hg(II)–GGAMSAASP14.

TGA Analyses

TGA thermogram of GG (Figure a) envisaged the removal of loosely bound moisture and volatile components within 50–130 °C, followed by the commencement of the major degradation at 230 °C.[60] In fact, such a rapid degradation continued up to 320 °C resulting in the mass loss of 49 wt %. In succession, slightly delayed thermal degradation caused a mass loss of 71 wt % at 400 °C.[60] Indeed, being a polysaccharide, GG went through a complicated thermal degradation process, constituting initial depolymerization via random chain scission followed by molecular rearrangements, resulting in the formation of an almost negligible residue at 700 °C.

Figure 4

TGA of (a) GG, AMSAASP14/18, GGAMSAASP14/18, and MV–/Hg(II)–/Hg(II)–MV–GGAMSAASP14/18; DSC of (b) GG, AMSAASP14/18, GGAMSAASP14/18, (c) GGAMSAASP14, MV–/Hg(II)–/Hg(II)–MV–GGAMSAASP14, and (d) GGAMSAASP18, and MV–/Hg(II)–/Hg(II)–MV–GGAMSAASP18.

Followed by the completion of moisture loss and associated anhydride formation via dehydration of neighboring −COOH within 200 °C, the thermal decomposition of AMSAASP progressed through three successive temperature ranges: 200–350, 350–420, and 420–500 °C (Figure a). In fact, the first stage of decomposition was associated with the conversion of neighboring amides of AMSAASP into imides, followed by the degradation of −COO– and imides, and finally, the third stage was attributed to the formation of paraffinic materials via scission of the AMSAASP backbone.[61] However, significantly high residues of 36.14/33.90 wt % were ascribed to the formation of thermoresistant inorganic salts and carbon of AMSAASP14/18. Interestingly, a marginally lower amount of residue for AMSAASP18 than AMSAASP14 was associated with the formation of a relatively higher extent of anhydride via dehydration of the frequently available neighbouring −COOH in AMSAASP18 at the initial stage, followed by decarboxylation of anhydrides in the higher temperature range.

Grafting of GG with AMSAASPs could be possible via attachment of GG with either SA or AM. Therefore, the combined thermal decomposition behavior of both GGSA and GGAM should be manifested during the thermal decomposition of both the GGAMSAASPs. In fact, GGAMSAASP18 showed better thermal resistance up to 420 °C, as compared to both the AMSAASPs. Thereafter, the degradation profile of both GGAMSAASP18 and AMSAASP18 remained almost similar up to 600 °C (Figure a), and thereafter, AMSAASP18 was found to be more thermoresistant than GGAMSAASP18, leading to a significantly higher residue for AMSAASP18. Almost similar thermal degradation contours were observed within 600–700 °C in the comparative plots of both GGAMSAASP14 and AMSAASP14. Indeed, relatively lower residues in both GGAMSAASPs were rationalized by the inclusion of negligible residue-forming GG within AMSAASPs (Figure a). Thus, during the first two stages of decomposition, the slightly better thermal resistance of GGAMSAASP14 over AMSAASP14 was attributed to the restricted conversion of amides to imides by the incorporation of grafted GG chains in between two closely spaced amide side chains. In a similar fashion, the grafted GG chains resisted anhydride formation via arresting the dehydration of neighbouring −COOH at the initial stage and associated decarboxylation of anhydrides in the later stages by partially affecting the close proximity between the adjacent −COOH. Interestingly, the overall thermal stability of GGAMSAASP14 was considerably inferior to that of GGAMSAASP18. Indeed, within 340–500 °C, significant deterioration in the thermal resistance of GGAMSAASP14 was observed as compared to that of GGAMSAASP18. Nevertheless, GGAMSAASP14 showed superior thermal resistance up to 260 °C as compared to both the AMSAASPs. It was believed that grafting of GG in GGAMSAASP14 was relatively unstable and less intimate in comparison to that observed in GGAMSAASP18. Such phenomena could also be correlated with the relatively enhanced thermal stability of GGAA over GGAM.[60] Accordingly, GGAMSAASP18 reflected a superior thermal resistance due to relatively higher availability of GGSA moieties in the network. Moreover, because M(II) adsorption was conducted at pHi > pHPZC, an enormous portion of −COOH was converted to −COO– and thus drastically reduced the extent of moisture loss within 30–200 °C, via anhydride formation from the neighboring −COOH of Hg(II)–GGAMSAASPs than that of GGAMSAASPs (Figure a). In this regard, a substantial reduction in the moisture loss was observed for Hg(II)–GGAMSAASP18 by the relative enhancement of −COO–, generated via ionization of the abundant −COOH in GGAMSAASP18.

In fact, the drastic degradation of both the Hg(II)–GGAMSAASPs was observed beyond 290 °C, and such a steep downfall continued up to 390 °C resulting in 50 wt % mass loss within 290–390 °C. Invariably, the mass loss was even more rapid in Hg(II)–GGAMSAASP14, containing greater proportion of amides because of the relatively enhanced preferential interaction of Hg(II) with N-donors. Such a phenomenon was associated mostly with the loss of amides in the first stage of thermal decomposition, within 200–350 °C, followed by the destruction of imides from 350 to 420 °C, resulting in the loss of N-donors, such as amides and imides, and thus facilitating the loss of mercury vapor at 350 °C. However, beyond 390 °C, the thermal decomposition was noted to be deaccelerated, as the amide destruction process was almost completed and the usual degradation process of polymer backbone commenced. In this regard, the decarboxylation of intermediate anhydrides appeared to impose little effect on the thermal decomposition of Hg(II)–GGAMSAASP18 throughout the entire degradation process, as unlike N-donors, O-donors possessed little tendency to interact with Hg(II). However, considerable extent of residues in Hg(II)–GGAMSAASPs, though lesser than GGAMSAASPs, were attributed to the prevalence of thermally stable inorganic components of AMSAASPs present within Hg(II)–GGAMSAASPs. In fact, the residue contents of various hydrogels were in the following order: GGAMSAASP18 (29.90 wt %) > GGAMSAASP14 (24.40 wt %) > Hg(II)–GGAMSAASP18 (20.45 wt %) > Hg(II)–GGAMSAASP14 (9.84 wt %). However, the relatively lower residue for Hg(II)–GGAMSAASPs than GGAMSAASPs was ascribed to the gradual replacement of Na+ by Hg(II) during adsorption and rapid evaporation of adsorbed mercury from Hg(II)–GGAMSAASPs. Notably, the difference between the amount of residues for Hg(II)–GGAMSAASPs (i.e., 20.45 – 9.84 = 10.61 wt %) was much higher than the corresponding difference between GGAMSAASPs (i.e., 29.9 – 24.4 = 5.5 wt %). One possible reason behind such an increase in the residue difference could be the higher Hg(II) adsorption by the amide-rich GGAMSAASP14 because Hg(II) preferably interacted with amides to produce thermolabile Hg–N covalent bonds and >N–Hg–N< cross-links, as envisaged earlier from FTIR analyses.

In contrast to Hg(II)–GGAMSAASPs, MV–GGAMSAASPs demonstrated superior thermal resistance along with higher residues than the respective GGAMSAASPs (Figure a). This phenomenon could be accounted for the higher individual thermal stability of MV over mercury along with relative lowering in the thermal degradation rate of MV up to 400 °C. Moreover, being flat and bulky, MV molecules could be involved in stacking interactions among themselves and with GGAMSAASPs, along with the usual formation of ionic bonds, leading to the enhanced thermal stability of MV–GGAMSAASPs.

Interestingly, both of the Hg(II)–MV–GGAMSAASPs showed inferior thermal resistances within 30–320 °C, as compared to the respective MV–/Hg(II)–GGAMSAASPs (Figure a).Because thermal stability at a lower temperature region was governed mainly by the physical interactions, such bonding among GGAMSAASPs and each of Hg(II)–MV was affected by the mutual interactions between individual adsorbates. In fact, thermal stabilities of both the Hg(II)–MV–GGAMSAASPs were noted to be improved within 320–600 °C and became even better than Hg(II)–GGAMSAASPs. However, thermal resistances of both Hg(II)–MV–GGAMSAASPs were significantly poor than the respective MV–GGAMSAASPs. Such a phenomenon also indicated a stronger mutual interaction between MV, Hg(II), and GGAMSAASPs within Hg(II)–MV–GGAMSAASPs, which arrested the thermal degradation via resisting the evaporation of mercury. Beyond 600 °C, the thermal degradations of both Hg(II)–MV–GGAMSAASPs were accelerated, and the extent of residues became almost similar to those of Hg(II)–GGAMSAASPs.

DSC Analyses

In the DSC thermogram of GG (Figure b), the endothermic transitions at 76 and 239 °C were ascribed to the loss of moisture and the cleavage of gal and mann units from the GG backbone, respectively.[62] Interestingly, the wide decomposition peak at 239 °C was considerably lower than the peaks at 253 and 248 °C,[62] which could be attributed to the shorter chain length and hence the lesser average molecular weight of GG than that of the average (i.e., 889 742.06 Da). Moreover, the appearance of the shoulder peak at 296 °C was also in agreement with the reported work.[62]

Interestingly, the DSC thermogram of AMSAASP18 (Figure b) exhibited a sharp endothermic peak at 159 °C, whereas the nature of initial thermal transition for AMSAASP14 was relatively broad and shifted to a higher temperature at 186 °C. As compared to AMSAASP18, relatively restricted endothermic transition in AMSAASP14 could be attributed to lesser anhydride formation from relatively fewer number of neighbouring −COOH in AMSAASP14. In fact, the formation of anhydride and the associated endothermic peak within 150–300 °C was feasible only when the prevalence of −COOH was more predominant than −COO–.[2] Indeed, considerably higher thermal stability of AMSAASP14 over AMSAASP18 was reflected from the occurrence of endothermic transitions at relatively higher temperatures of 242, 344, and 433 °C for AMSAASP14 than the respective peaks at 237, 344, and 431 °C for AMSAASP18. Notably, the endothermic peaks at 242 and 237 °C for AMSAASP14 and AMSAASP18, respectively, were attributed to the conversion of amides into imides, whereas the breakdown of imides and anhydrides was predominant at the higher temperature region. As established from the NMR and FTIR analyses, AMSAASP14 was constituted of numerous −CH2–NH–(C=O)– moieties produced via free-radical interactions between SA and •NH–(C=O)– during the synthesis of AMSAASP14, and such moieties eventually provided marginal improvement in the thermal stability of AMSAASP14 over AMSAASP18 throughout the entire temperature range. In fact, the enhanced conversion of −(C=O)–NH2 to −CH2–NH–(C=O)– was believed to be responsible behind the restricted degradation of amides to imides, leading to the overall retardation in the decomposition of AMSAASP14 in later stages. Eventually, such marginally improved thermal stability of AMSAASP14 over AMSAASP18 generated a higher residue for AMSAASP14 in the TGA analyses.

Grafting of GG brought about several changes in the DSC thermograms of both the AMSAASPs. For instance, the prevalence of two sharp peaks at 60 and 240 °C in GGAMSAASP14 could be related to the GG-specific endothermic transitions at 76 and 239 °C, respectively. Moreover, the broad peak at 181 °C in GGAMSAASP14 could be correlated with the similar endothermic transitions of AMSAASP14 at 186 °C. However, in the DSC thermogram of GGAMSAASP18, an extensively broad peak centered at 130 °C was generated, which was related to both GG and AMSAASP18-specific thermal transitions. In this context, the sharp endothermic peak at 159 °C for AMSAASP18 completely disappeared for GGAMSAASP18, suggesting extensive phase mixing that significantly affected the anhydride formation-driven heat absorption and moisture loss. Indeed, in comparison to GGAMSAASP14, the anhydride formation and associated heat absorption would mostly be affected in GGAMSAASP18, owing to the abrupt decrease in the neighboring −COOH via incorporation of grafted GG chains in between the adjacent −COOH. Interestingly, the occurrence of two broad and intense peaks at 230 and 352 °C in the thermogram of GGAMSAASP14, absent in both AMSAASP14 and GGAMSAASP18, indicated the enhanced contribution of frequently available thermo-labile GGAM segments in destabilizing the GGAMSAASP14 network. It was believed that the relatively easier breakdown of GGAM, as compared to GGSA, might accelerate the decomposition cascade via cyclization of amides as well as easier decomposition of thermally detached GG chains from the numerous thermo-labile GGAM segments of GGAMSAASP14.

The appearance of two new exothermic peaks at 194/202 °C (Figure c,d) in the DSC thermogram of Hg(II)–GGAMSAASP14/18 indicated the possible displacement reaction between the adsorbed HgCl2/Hg(OH)2 and Al-pan.[63] In fact, the deposition of Hg(OH)2 on both the Hg(II)–GGAMSAASPs had been established from the respective FTIR findings (Figure b). As compared to Hg(II)–GGAMSAASP18, the more intense exothermic peak of Hg(II)–GGAMSAASP14 could be ascribed to the enhanced displacement reaction between the higher number of adsorbed Hg(II) in Hg(II)–GGAMSAASP14 and Al-pan. Moreover, the adsorption of Hg(II) onto GGAMSAASP14 resulted in the complete disappearance of broad and intense endothermic peaks at 230 and 352 °C. In fact, the prevalence of thermally stable >N–Hg–N< cross-links, produced via interactions between amides and adsorbed Hg(II) in the bulk and on the surface of Hg(II)–GGAMSAASP14, as established earlier from NMR and FTIR analyses, resisted the thermal degradation of amides because free amides were rare in Hg(II)–GGAMSAASP14. Moreover, the appearance of sharp endothermic peaks at 223 °C for Hg(II)–GGAMSAASP14 could be compared to a similar peak for the chitosan–Hg(II) complex, actuated via complexation between N-donors and Hg(II).[64] However, fewer availability of >N–Hg–N< cross-links and accordingly greater extent of free amides in the bulk of Hg(II)–GGAMSAASP18 were also substantiated from the occurrence of a relatively broad endothermic peak at 288 °C in Hg(II)–GGAMSAASP18 than the low intense endothermic peaks at relatively higher temperatures of 296 and 317 °C for Hg(II)–GGAMSAASP14.

The DSC thermograms of MV–GGAMSAASPs envisaged almost complete disappearance of the broad endothermic peaks within 150–300 °C, especially in MV–GGAMSAASP18. This phenomenon could be attributed to the substantial conversion of −COOH into −COO– during MV adsorption onto GGAMSAASPs at alkaline pHi, especially in MV–GGAMSAASP18, leading to the minimization of anhydride formation possibility via dehydration of fewer number of available −COOH. Because GGAMSAASP18 contained a higher amount of −COOH, the impact of alkaline pH-driven conversion of −COOH into −COO– would be the maximum in GGAMSAASP18. Altogether, MV–GGAMSAASPs showed greater thermal stabilities as compared to both GGAMSAASPs and Hg(II)–GGAMSAASPs, as realized from the prevalence of a relatively low intense endothermic peak at a higher temperature in the thermograms of MV–GGAMSAASPs. Such phenomena were also apprehended earlier during TGA analyses, and the increased thermal resistance of MV–GGAMSAASPs could be correlated with the cross-linking ability of MV molecules via multipoint attachment within the hydrogel matrices.

In both the Hg(II)–MV–GGAMSAASPs, the disappearance of the exothermic peaks at 194 and 202 °C (Figure c,d) indicated the absence of free Hg(II) as Hg(OH)2/HgCl2 at Hg(II)–MV–GGAMSAASP surfaces, established earlier from the FTIR results. In fact, the obsolescence of free Hg(II) as HgCl2/Hg(OH)2 could only be made possible via formation of a stable Hg(II)–MV complex, and this was unable to undergo amalgamation with Al-pan. In fact, Hg(II)–MV interaction resulted in impaired stability of both the Hg(II)–MV–GGAMSAASPs than MV–GGAMSAASPs, preferably in the lower temperature region, together with the reappearance of a relatively less intense and broad endothermic peak at 234 °C in Hg(II)–MV–GGAMSAASP14, similar to that of GGAMSAASP14. Similarly, a small endothermic peak at 245 °C in MV–GGAMSAASP18 was shifted to 225 °C in Hg(II)–MV–GGAMSAASP18, which suggested easier conversion of free amides to imides. Indeed, such amides would be generated due to a mutual Hg(II)–MV interaction. Again, such a mutual interaction was also responsible for the possible shifting of the endothermic peak from 340 °C in MV–GGAMSAASP18 to 334 °C in Hg(II)–MV–GGAMSAASP18. Moreover, the arrival of multiple endothermic peaks within 340–371 °C in Hg(II)–MV–GGAMSAASP14 suggested variegated interactions among MV, Hg(II), and GGAMSAASP14, which were not found in MV–/Hg(II)–GGAMSAASP14.

XRD Analyses

The intimate phase mixing and hence dislocation of regular spatial arrangements through the grafting of GG onto terpolymers in both the GGAMSAASPs was comprehended via complete disappearance of characteristic GG peaks at 20.44°, 40.57°, and 70.66° (Figure a),[62] along with the simultaneous absence of the terpolymeric peaks. However, relatively higher crystallinity of AMSAASP14 over AMSAASP18 was realized from the appearance of a couple of peaks at 23.21° and 29.64° in AMSAASP14, as compared to the comparatively broader and lesser intense peak at 32.65° for AMSAASP18 (Figure a). In fact, such an enhanced crystallinity of AMSAASP14 over AMSAASP18 was also observed in the respective GGAMSAASPs through the appearance of a relatively ordered structure in GGAMSAASP14 than GGAMSAASP18.

Figure 5

XRD of (a) GG, AMSAASP14/18, GGAMSAASP14/18, (b) GGAMSAASP14, MV–/Hg(II)–/Hg(II)–MV–GGAMSAASP14, (c) GGAMSAASP18, and MV–/Hg(II)–/Hg(II)–MV–GGAMSAASP18.

Indeed, some new peaks were generated in MV–GGAMSAASPs, which were not found in GGAMSAASPs. In fact, the XRD spectrum of MV–GGAMSAASP14 showed two new peaks at 36.55° and 44.96°, along with the symbolic peak at 30.74° for GGAMSAASP14 (Figure b). Of these, the peak at 36.55° was ascribed to the shortest distance of 2.46 Å between every two carbon atoms, allocated at the meta-positions in the benzene rings of MV. However, the new prominent peak at 45.09°, having d-spacing = 2.01 Å at n = 1, correlated with the spacing of liquid benzene (i.e., 2.00 Å). Similarly, the arrival of a new peak at 22.96° in MV–GGAMSAASP18 (Figure c) might be attributed to the interatomic distances of 3.87 Å in the planar aromatic structure of MV+, attached on the GGAMSAASP18 surface. In this context, a similar interatomic distance of 3.83 Å was also substantiated by the radial distance of 3.75 Å of a single layer of graphite comprising of interconnected benzene rings.[65]

The diffractogram of Hg(II)–GGAMSAASP14 possessed several new peaks, of which the peaks at 21.79° and 43.85° could be related to the characteristic peaks of HgCl2[66] that confirmed the prevalence of HgCl2 crystals on the surface of Hg(II)–GGAMSAASP14. On the contrary, such a kind of appearance was not realized on the surface of Hg(II)–GGAMSAASP18. In fact, deposition of lesser amount of HgCl2 on the Hg(II)–GGAMSAASP18 surface was also substantiated from the appearance of a less intense exothermic peak in the respective DSC thermogram (Figure c,d). However, such HgCl2 peaks could not be detected in Hg(II)–MV–GGAMSAASP, corroborating the mutual interaction of Hg(II) and MV at the Hg(II)–MV–GGAMSAASP interfaces, as demonstrated earlier in the respective FTIR spectra (Figure b). Nevertheless, despite the Hg(II)–MV interaction, the characteristic MV+ peak at 23.01° prevailed in both the Hg(II)–MV–GGAMSAASPs. However, the appearance of an extra symbolic MV+ peak at 45.10° in Hg(II)–MV–GGAMSAASP18 indicated higher population of MV on the GGAMSAASP18 surface. In a nutshell, individual adsorption of Hg(II) resulted in a significant deterioration of the crystallinity of both the GGAMSAASP surfaces, whereas individual MV adsorption enhanced the crystallinity. However, the combined Hg(II)–MV adsorption ensued intermediate level of crystallinity, which reemphasized the loss of an ordered structure via mutual Hg(II)–MV interactions.

FESEM and EDX Analyses

From the FESEM microphotographs of AMSAASP14/18, distinct phase boundaries were observed (Figure a,b). These relatively heterogeneous surfaces showed several discontinuous cracks. However, such surfaces of GGAMSAASPs were devoid of such distinct phase boundaries and appeared to be smooth, well-organized, and compact as a result of finer intermixing through the grafting of GG within the network of AMSAASPs (Figure c,d). Nevertheless, an uneven and heterogeneous surface appeared because of deposition of Hg(II) onto GGAMSAASPs (Figure e,f). In fact, the maximum Hg(II) deposition ascertained from the ACs of GGAMSAASP14, as compared to that of GGAMSAASP18, was also comprehended through the appearance of a fully heterogeneous surface in FESEM and attainment of the maximum Hg(II) deposition from EDX analyses. In this context, formation of covalent, coordinate, and/or ionic bonds between GGAMSAASPs and Hg(II) were eventually realized from the appearance of variegated Hg(II) peaks in the EDX spectrum (inset of Figure e,f).

Figure 6

FESEM microphotographs of (a) AMSAASP14, (b) AMSAASP18, (c) GGAMSAASP14, (d) GGAMSAASP18, (e) Hg(II)–GGAMSAASP14, and (f) Hg(II)–GGAMSAASP18; EDX spectrum of (inset of e/f) Hg(II)–GGAMSAASP14/18.

RSM-Optimized Hydrogel Synthesis

Phase-1: Screening of Synthetic Parameters To Identify the Key Variables

In the phase-1 analysis, a set of 19 experiments were conducted for the screening of the most significant variables based on the resolution-IV design (Table ). The maximum/minimum levels chosen for such analysis were 6.25/25.00, 0.20/0.60, 2.00/4.00, 1.00/5.00, and 4.00/10.00 wt % for A, B, C, D, and E, respectively, whereas the maximum/minimum ESR values were varied within 4.20–16.80. However, only three variables, that is, A, D, and E, were observed to cross the Bonferroni limit of 4.3817 in the Pareto chart (Figure S3) and hence considered for further CCD analysis. In addition, the Pareto chart also recognized BE as the only 2FI term to impart adequate effect on ESR. Therefore, these four terms, that is, three 1 factor and one 2FI, were chosen for constituting a regression model whose applicability was evaluated from the following second-order polynomial equation (eq ).

Table 2

Resolution-IV Design for Screening of Important Process Variables in Phase-I

run no.	amount of AM (wt %)	total amount of initiator (wt %)	amount of GG (wt %)	amount of cross-linker (wt %)	pHi (−)	ESR (−)
1	6.25	0.20	2.00	5.00	10.00	13.21
2	25.00	0.20	2.00	1.00	4.00	5.50
3	6.25	0.60	2.00	1.00	10.00	16.80
4	25.00	0.60	2.00	5.00	4.00	4.20
5	6.25	0.20	4.00	5.00	4.00	5.53
6	25.00	0.20	4.00	1.00	10.00	16.10
7	6.25	0.60	4.00	1.00	4.00	10.80
8	25.00	0.60	4.00	5.00	10.00	9.83
9	15.625	0.40	3.00	3.00	7.00	10.00
10	15.625	0.40	3.00	3.00	7.00	10.00
11	15.625	0.40	3.00	3.00	7.00	10.00

In fact, the chosen model was found to be sufficiently significant owing to very high adj. R2 value (i.e., 0.9745), closeness of adj./pred. R2 values (i.e., 0.9745/0.8995), and a very low p-value (<0.0001) (Table ).

Table 3

ANOVA Statistics in Phase-I

source	sum of squares	degrees of freedom	mean square	F value	p-value
model	165.44	4	41.36	87.11	<0.0001*
amount of AM (A)	14.34	1	14.34	30.20	0.0027*
amount of cross-linker (D)	33.74	1	33.74	71.07	0.0004*
pHi (E)	111.83	1	111.83	235.54	<0.0001*
BE	5.53	1	5.53	11.64	0.0190*
curvature	0.13	1	0.13	0.28	0.6202
residual	2.37	5	0.47
lack of fit	2.37	3	0.79
pure error	0.00	2	0.00
cor. total	167.94
std dev.	0.69	R2	0.9859
mean	10.18	adj. R2	0.9745
CV %	6.77	pred. R2	0.8995
PRESS	16.88	adeq. precision	24.7600

Phase-2: CCD Optimization of the Three Most Significant Process Variables

CCD, a standard RSM design, was adopted to optimize the three most significant variables of synthesis, that is, A, D and E, to understand the individual and/or interactive effects on the response, that is, ESR. In this context, CCD is a combination of 2, 2n, and nc numbers of factorial design, axial, and central points, respectively. The predesigned experiments were performed randomly for minimizing the effect of uncontrolled factors.[67] The results were examined and correlated with input variables using the second-order empirical polynomial eq .Here, Y, β0, β, β, and β represent the predicted response, constant, linear, quadratic, and interaction coefficients, respectively. In this regard, the analysis of variance (ANOVA) table was studied to justify the significance and adequacy of the predicted model by taking input variables within 6.25–25.00 wt %, 1.00–6.36 wt %, and 4–13.50 wt %. The input variables, both in coded and real, and the response, as generated by the software, are listed in Table S1. In this context, ANOVA table was also incorporated to study the individual and/or synergistic effects of these variables via estimating R2 values to correlate the experimental and predicted data at 95% confidence level (Table S2).[67] In addition, the sequential model sum of squares (type I) and the model summary statistics tests were also conducted to rationalize the interactive effects of A, D, and E on ESR via linear, 2FI, quadratic, and cubic models. However, the quadratic model was the best for determining the experimental data because of the highest R2 values (i.e., 0.9948) and the closest resemblance of adj. and pred. R2 values (0.9901 and 0.9597). The fitment of experimental data to eq resulted in eq .

The response surface plots, showing the interactive effects of AD, AE, and DE, are given in Figure . In fact, the overall optimization was carried out to obtain the perfect conditions for synthesizing a sustainable hydrogel having the maximum ESR. In this regard, it is also essential to mention that the hydrogel having higher swellability in water corresponds to a higher AC. Thus, in the present study, the GGAMSAASP possessing the maximum ESR has been rationally selected for the adsorptive removal of MV and Hg(II). Thus, A, D, and E “in range” and ESR “maximize” were considered in the numerical optimization section. However, the optimum conditions for attaining the maximum ESR of 21.41 were 8.98, 1.00, and 10.00 wt % for A, D, and E, respectively.

Figure 7

Swelling reversibility of (a) GGAMSAASP18/14 and AMSAASP18/14; pHPZC of (b) GGAMSAASP18/14 and swelling study of (c/d) GGAMSAASP18/14 at different pHi values; 3D response surface plots of ESR (−) vs (e) amount of MBA (wt %) and amount of AM (wt %), (f) amount of MBA (wt %) and pHi (−), and (g) amount of MBA (wt %) and pHi (−).

Adsorption Isotherm Study

The nature of adsorbate–adsorbent interaction during isothermal adsorption can be rationalized through fitment of experimental data to Langmuir, Freundlich, and Sips (eqs S1–S4) isotherm models (Figure , S4). In the present case, adsorption isotherm studies were conducted by taking 0.025 g of GGAMSAASPs and 5–30 ppm dye solutions at pHi = 7/9 for Hg(II)–MV within 293–323 K. However, the Langmuir model was found to be the best for explaining the experimental equilibrium data of Hg(II)–GGAMSAASP14/18 (Table ) in the entire temperature and concentration ranges (Figure a and inset). According to the presumptions of the Langmuir model, monolayer adsorption of Hg(II) on the structurally consistent hydrogel surface could be interpreted. From Table , higher qmax of Hg(II) was observed for GGAMSAASP14, which indicated greater affinity of Hg(II) on GGAMSAASP14. This phenomenon could be explained by considering higher covalent bonding affinity of Hg(II) with the N-center. In fact, higher affinity of Hg(II) to produce O=C–NH–Hg(II)–HN–C=O had already been recognized from the respective XPS analysis via considerable shifting of N 1s BE from 399.95 to 399.35 eV. This phenomenon also indicated the prevalent chemisorption for Hg(II) on GGAMSAASPs. However, for both GGAMSAASPs, qmax was found to increase with increasing temperature. Again, separation factor (RL) (eq ) was found to lie within 0–1, which indicated the prevalence of favorable adsorption.

Figure 8

Langmuir fitting for (a/inset of a) Hg(II)–GGAMSAASP14/18; Sips and Langmuir fitting for (b) MV–GGAMSAASP18 and (c) MV–GGAMSAASP14, respectively; (inset of b/c) full scan MV adsorption spectrum for GGAMSAASP18/14 at pHi = 9.

Table 4

Adsorption Isotherm and Kinetics Parameters

temperature (K)
models/parameters	293	303	313	323
Hg(II)
Langmuir (GGAMSAASP18)
qmax (mg g–1)/pHi/C0 (ppm)	37.51/7/5–30	40.95/7/5–30	42.99/7/5–30	43.01/7/5–30
kL (L mg–1)	4.23	2.81	2.42	2.40
R2	0.9963	0.9993	0.9995	0.9995
F	3318.13	15 968.97	21 432.52	21 326.51
χ2	3.89	0.14	0.14	0.14
Langmuir (GGAMSAASP14)
qmax (mg g–1)/pHi/C0 (ppm)	48.38/7/5–30	49.12/7/5–30	49.45/7/5–30	50.65/7/5–30
kL (L mg–1)	1.63	1.65	1.73	1.73
R2	0.9888	0.9901	0.9876	0.9874
F	989.96	1104.94	878.76	859.28
χ2	3.22	2.93	3.75	3.92
Freundlich (GGAMSAASP18)
kF (mg g–1 L1/n mg–1/n)	25.09	25.73	26.03	27.86
n/pHi/C0 (ppm)	5.13/7/5–30	4.55/7/5–30	4.08/7/5–30	3.11/7/5–30
R2	0.9309	0.9493	0.9385	0.9458
F	176.99	141.35	187.23	197.94
χ2	14.39	19.36	15.41	16.76
Freundlich (GGAMSAASP14)
kF (mg g–1 L1/n mg–1/n)	26.89	27.24	27.78	28.45
n/pHi/C0 (ppm)	3.39/7/5–30	3.32/7/5–30	3.28/7/5–30	3.19/7/5–30
R2	0.9612	0.9659	0.9701	0.9717
F	284.13	320.82	363.77	380.58
χ2	11.14	10.03	9.01	8.82
Pseudo-Second Order (GGAMSAASP18)
qe,cal (mg g–1)/pHi/C0 (ppm)	38.88/7/30	39.89/7/30	42.11/7/30	41.97/7/30
qe,exp (mg g–1)	37.93 ± 1.14	39.36 ± 1.23	41.65 ± 1.45	46.98 ± 2.07
k2 (g mg–1 min–1)	0.0038	0.0046	0.0064	0.0081
R2	0.9899	0.9967	0.9911	0.9932
F	6796.22	16 861.21	7083.34	9667.84
χ2	1.14	0.46	1.28	0.95
Pseudo-Second Order (GGAMSAASP14)
qe,cal (mg g–1)/pHi/C0 (ppm)	46.01/7/30	46.68/7/30	47.25/7/30	47.96/7/30
qe,exp (mg g–1)	45.21 ± 1.37	45.89 ± 1.39	46.54 ± 1.44	47.36 ± 2.01
k2 (g mg–1 min–1)	0.0045	0.0052	0.0079	0.0097
R2	0.9993	0.9983	0.9965	0.9898
F	75 278.79	33 369.79	16 089.74	5798.32
χ2	0.13	0.31	0.68	1.93

The equilibrium adsorption of MV followed the Sips/Langmuir isotherm models for GGAMSAASP18/14 (Figure b,c). However, relatively higher qmax was observed for GGAMSAASP18 (Table ). Therefore, the enhancement of −COO– facilitated the formation of the ionic interaction between −NHMe+/–NMe2+ of MV and −COO– on the surface of GGAMSAASP18 and thus the attainment of higher AC, evidenced from the respective FTIR and TGA analyses. Again, kS and kL were found to increase with the enhancement of temperature.[2]

Adsorption Thermodynamics Study

Thermodynamic spontaneity of chemisorption was ascertained from the measurements of ΔG0 values using eq .Here, kd, known as the distribution coefficient, can be defined by eq .

In the entire concentration and temperature ranges of the experiment (Table ), ΔG0 values were negative, which indicated thermal spontaneity of chemisorption.[23] From Table , higher −ΔG0 values were observed for MV–GGAMSAASP18 than for MV–GGAMSAASP14, as also indicated from the respective qmax values. In fact, the values of −ΔG0 were found to increase with the rise in temperature from 293 to 323 K, reflecting a more favorable adsorption at a higher temperature. Conversely, for Hg(II) adsorption, −ΔG0 values were found to be higher for GGAMSAASP14 than for GGAMSAASP18, indicating higher adsorption feasibility for GGAMSAASP14. The prevalence of larger number of N-centers in GGAMSAASP14 alleviated Hg(II) adsorption, whereas the population of a larger amount of −COO– facilitated the adsorption of MV. Additionally, the exothermic nature of adsorption was rationalized by the negative values of ΔH0 (Figure S5e–h), whereas +ΔS0 values (eq ) suggested fair affinity of MV/Hg(II) onto GGAMSAASP14/18 along with the decrease in randomness at the solid–solution interface during adsorption.

Table 5

Adsorption Thermodynamics Parameters

concentration (ppm) of Hg(II)/MV/temperature (K)	–ΔG0 (kJ mol–1) of Hg(II)/MV for GGAMSAASP18(GGAMSAASP14)	–ΔH0 (kJ mol–1) of Hg(II)/MV for GGAMSAASP18(GGAMSAASP14)	ΔS0 (J mol–1 K–1) of Hg(II)/MV for GGAMSAASP18(GGAMSAASP14)
05/05/293	11.42(11.16)/7.32(12.60)	14.23(−1.01)/6.89(−32.72)	–9.79(41.47)/3.26(141.58)
05/05/303	11.07(11.54)/7.19(11.68)
05/05/313	11.33(11.98)/6.96(11.08)
05/05/323	11.01(12.39)/7.21(8.17)
10/10/293	10.81(9.42)/8.54(9.88)	13.16(−7.27)/6.32(−7.66)	–8.41(56.84)/6.82(54.35)
10/10/303	10.41(9.87)/8.43(9.36)
10/10/313	10.62(10.58)/8.37(8.81)
10/10/323	10.47(11.08)/8.33(8.25)
15/15/293	8.64(8.03)/8.51(8.09)	–0.44(−3.30)/5.27(1.26)	31.01(38.64)/9.97(21.20)
15/15/303	8.95(8.42)/8.36(7.91)
15/15/313	9.26(8.74)/8.28(7.70)
15/15/323	9.58(9.21)/8.20(7.46)
20/20/293	6.06(7.45)/8.54(7.70)	–15.57(−0.60)/6.32(−5.31)	74.02(27.48)/6.82(40.36)
20/20/303	7.08(7.72)/8.43(7.35)
20/20/313	7.31(7.99)/8.37(6.94)
20/20/323	8.46(8.28)/8.34(6.50)
25/25/293	3.89(5.72)/8.04(6.47)	1.31(−6.96)/4.60(−5.92)	14.10(43.11)/10.78(38.40)
25/25/303	4.75(6.05)/8.03(6.11)
25/25/313	5.26(6.47)/7.89(5.75)
25/25/323	6.92(7.02)/7.73(5.31)
30/30/293	3.01(4.41)/7.29(5.44)	–18.70(−5.29)/–5.23(−9.57)	73.38(33.04)/38.75(46.47)
30/30/303	3.34(4.72)/6.86(4.94)
30/30/313	3.94(5.03)/6.54(4.56)
30/30/323	5.31(5.41)/6.10(4.02)

Adsorption Kinetics Study

The fitment of pseudo-second-order kinetics (eq ), ascertained by the higher R2, F, and the closest proximity of qe,exp and qe,cal (Table ), for MV and Hg(II) indicated the prevalence of chemisorption in the entire temperature (Figure S5a–d) and concentration ranges of the experiment (Figure a–d). In fact, the existence of such chemisorption was also apprehended by the increasing trend of the pseudo-second-order rate constant, k2 (g mg–1 min–1), with increasing temperature for GGAMSAASPs. The higher k2 for Hg(II)–GGAMSAASP14 indicated the preferential Hg–N bond formation with the N-donor ligand of AM, also harmonized from the XPS and FTIR analyses (Figures b and 3e). By contrast, the enhancement of k2 values for MV–GGAMSAASP18 was ascribed to the higher number of −COO– in GGAMSAASP18.

Figure 9

Pseudo-second-order kinetics plots for (a/b) Hg(II)–GGAMSAASP18/14 (pHi = 7, T = 303 K, and adsorbent dose = 0.05 g L–1); pseudo-second-order kinetics plots for (c/d) MV–GGAMSAASP18/14 (pHi = 9, T = 303 K, and adsorbent dose = 0.05 g L–1); Weber and Morris fitting for (e/f) Hg(II)–/MV–GGAMSAASP14/18.

Study of Diffusion Mechanism

The plots of q vs t0.5 (eq ) of Hg(II)–MV for GGAMSAASPs were found to possess three linear portions of variable slopes and intercepts (Figure e,f), of which slopes of the second segments determined the values of kip (Table ).

In fact, the coexistence of bulk, film, and pore/intraparticle diffusions for Hg(II)–MV could comprehensively be rationalized by the appearance of such multilinear segments. However, the prevalence of intraparticle diffusion was also confirmed by the fitment of experimental data to the following nonlinear form of the Boyd model (eq ).Here, q/qe (mg g–1) and B represent ACs at time t/equilibrium and Boyd parameter, respectively. However, Boyd plots for Hg(II)–MV were found to be linear (Figure S5k,l) and started from the origin that reflected the prevalence of pore diffusion. Interestingly, the kip values of Hg(II) was found to be higher in GGAMSAASP14 than in GGAMSAASP18.

Effect of Temperature on Adsorption Kinetics

The effect of temperature on chemisorption kinetics was studied by taking 25 ppm Hg(II)–MV solutions at pHi = 7/9 and 0.025 g of GGAMSAASPs at 296, 303, 310, and 317 K. As all dyes followed pseudo-second-order kinetics, k2 at different temperatures could be interrelated by the following Arrhenius type eq .Here, k0 and Ea are the temperature independent factor (g mg–1 min–1) and activation energy of adsorption (kJ mol–1), respectively. Indeed, from the slope of the linearized ln k2 vs 1/T plot, Ea of adsorption can be evaluated (Figure S5i,j). However, Ea values for MV–Hg(II) were found to be 24.35/20.20 and 23.95/21.37 kJ mol–1 onto GGAMSAASP18 and GGAMSAASP14, respectively, indicating the prevalence of chemisorption in both the cases.[68]

Synergistic Adsorption of Hg(II) and MV by GGAMSAASPs

At the beginning of simultaneous Hg(II) and MV adsorption by the GGAMSAASPs, mutual interaction between Hg(II) and MV was reflected via appearance of a strong hypochromic effect in the visible spectrum of Hg(II) and MV than that of individual MV. In addition, a time-dependent hypsochromic effect was also observed as a consequence of the Hg(II)–MV interaction, resulting in a significant shift of λmax from 577.85 to 544.35 nm within 240 min. Mutual interaction between triphenyl methane dyes (TPMDs) and M(II) were reported earlier during quantifying M(II) in aqueous solution using TPMDs as indicators.[24] It was believed that the instantaneous hypochromic effect, together with the time-dependent hypsochromic effect, originated from the delocalization of electrons in MV+ by Hg(II), possibly via coordinate bond formation via electron donating −NHMe+/–NMe2+ moieties of MV to the vacant orbitals of Hg(II). Moreover, like crystal violet cations, MV+ should also exist as pyramidal and planar forms, with considerable difference between the λmax values of the two conformers. In fact, the λmax values for planar and pyramidal conformers of the crystal violet cation are 557 and 590 nm, respectively. Accordingly, λmax for the planar conformer of MV+ should be lower than the respective pyramidal conformer. The coordinate bonding between Hg(II) and −NMe2/–NHMe moieties of MV+ should be more favorable if the distance between the positively charged centers of the interacting cations are maximized to ensure minimum repulsion between Hg(II) and MV+ (Figure a,b). Invariably, the maximum separation between interacting Hg(II) and MV+ could be accomplished if MV+ exists as a planar form, wherein the positive charge is located at the terminal positions. Accordingly, in the presence of Hg(II), the rapid equilibrium between planar and pyramidal conformers should shift in favor of the planar form of MV+, possessing lower λmax than the pyramidal conformer.

Figure 10

Synergistic adsorption of (a/b) Hg(II)–MV–GGAMSAASP18/14; recyclability plots of (c) MV–GGAMSAASP14/18 and Hg(II)–GGAMSAASP14/18; photographic images of (d/inset of d) MV–GGAMSAASP18/GGAMSAASP18 and (e/f) desorption/adsorption kinetics of MV.

Mutual interaction between Hg(II)–MV was also realized from the significant change in the UV region of the UV–vis spectra of Hg(II) and MV. In the absence of Hg(II), the UV–vis spectra of MV demonstrated a peak at 299.30 nm (inset of Figure b,c). However, as Hg(II) was added to the MV solution, the MV specific peak at 299.35 nm was transformed into a shoulder positioned at 301.80 nm. Earlier, researchers claimed that significant interaction between aniline and Hg(II) would result in an appreciable red shift of characteristic azine-based receptor.[69] In the presence of Hg(II), shift of MV to a higher wavelength should also be an outcome of the mutual Hg(II)–MV interaction to produce at least a labile Hg(II)–MV complex.

Altogether, as compared to GGAMSAASP18, GGAMSAASP14 envisaged a relatively better combined adsorption of MV–Hg(II), realized from the superior hypochromic effect in the time-dependent visible spectra for GGAMSAASP14. Such phenomena could be related to the relatively greater availability of superficial −COO– on the GGAMSAASP18 surface that caused surface deposition of cationic adsorbates, thereby resisting further penetration of the adsorbates deep into the bulk. Indeed, as compared to GGAMSAASP18, relatively greater adsorption of combined adsorbates by GGAMSAASP14 was also realized from the greater residue difference between GGAMSAASP14 and Hg(II)–MV–GGAMSAASP14 (i.e., 24.34 – 11.52 = 12.82 wt %), as compared to the same between GGAMSAASP18 and Hg(II)–MV–GGAMSAASP18 (i.e., 29.86 – 21.91 = 7.95 wt %).

Desorption and Reusability

The recyclability of GGAMSAASPs was explored through the repetitive adsorption studies at pHi = 9/7 for MV–Hg(II), followed by the desorption of MV–/Hg(II)–GGAMSAASPs at 0.1 (N) NaCl/pHi = 2 (Figure c–f). Although, the desorption ratios of both the GGAMSAASPs decreased gradually in the successive cycles, the maximum desorptions of 95 and 85% were noted for MV–GGAMSAASP18/14 and Hg(II)–GGAMSAASP18/14, respectively, upto the fourth cycle, with the retention of the network. In fact, such results of desorption could rationally be correlated with the formation of physicochemical interactions between −NH2 of −CONH2, −COO– of GGAMSAASPs, and cationic N-centres of MV–Hg(II), as evidenced earlier by FTIR, NMR, and XPS analyses.

Comparison of the Results

Several micro/nanomaterial-based adsorbents, blends, homo/co/terpolymers, and IPN hydrogels for the adsorptive removal of MV and Hg(II) from aqueous solutions of varying initial concentrations (i.e., 1–3800 ppm), temperatures (i.e., 288–323 K), and pHi values have been provided (Table S3). In this context, it is important to note that the AC of any adsorbent not only depends on the initial concentration but also is a function of the temperature and pHi. From Table S3, it could be observed that the ACs of GGAMSAASPs were either closer to or much better than previously reported adsorbents, within the specified working range.

Conclusions

This report not only describes the synthesis of a novel sustainable IPN superadsorbent but also develops eventually a new pathway for the in situ green synthesis of GG-g-terpolymer without ex situ addition of a third monomer. This attractive alternative pathway for the synthesis of a new GGAMSAASP superadsorbent of alleviated physicochemical properties is confirmed through extensive 1H NMR, FTIR, XPS, TGA, and DSC analyses. The in situ allocation of −CO–NH–(CH2)2–COO– is established via 1H NMR, and the arrival of the O 1s shakeup satellite band in XPS spectra is corroborated by FTIR and thermal analyses. The grafting of GG into the AMSAASP terpolymer is confirmed by the appearance of −CH2–O–CH2– in the FTIR and NMR studies, supported by XRD and FESEM analyses. Relatively homogeneous distribution of Hg(II) in the entire amide-populated GGAMSAASP14 is inter-related to the preferential BB coordination with −COO–, along with greater surface deposition of insoluble Hg(OH)2 and HgCl2, confirmed by FTIR, DSC, TGA, and XRD studies, with relative ease of adsorption and difficulty during desorption of Hg(II) by the amide-populated GGAMSAASP14 because of the better relative availability of Hg–N bonds, such as >N–Hg–N< and −Hg–NH–, in Hg(II)–GGAMSAASP14, and envisaged from the higher k2 for Hg(II)–GGAMSAASP14 over Hg(II)–GGAMSAASP18. The binding of Hg(II) preferentially occurs on the surface of GGAMSAASP18, without Hg(II) in the core, generating the frequent M mode of coordination. In Hg(II)–MV binary adsorption, manifestation of the mutual Hg(II)–MV+ interaction infers the disappearance of insoluble Hg compounds in Hg(II)–MV–GGAMSAASPs, as evident from FTIR, TGA, DSC, and XRD and the combined hypsochromic–hypochromic shift in the UV–vis spectra. Altogether, owing to the rationally optimized combination of synthetic and NPs, imparting excellent physicochemical properties and performance characteristics in the chemisorption of MV and/or Hg(II), GGAMSAASPs show novelty (Table S3) in a kinetically fast green waste remediation process. This systematic design of GGAMSAASPs of excellent thermomechanical properties can also be attempted for drug delivery, tissue engineering, membrane-based separation, sensors, and self-healing materials. The redemption of this study is the introduction of this new pathway, which can also be attempted for synthesizing ASP-based terpolymers/IPNs, without orthodox ex situ addition of ASP or derivatives of ASP.