Introduction

Polyacrylics are a special class of polymers as far as their uses, and applications are of interest. Their nowadays ubiquity stems from the large spectrum of their properties which are on a par with the chemical reactivity of their functional groups, i.e., carboxylic, ester, amide, and nitrile. A wide range of their chemical modifications are, but the result of the proneness of these groups to chemical reactions, affording valuable materials with targeted properties. Poly(acrylic acid) (PAA), coined the name “carbopol” in pharmaceuticals, is known to be a super absorbent polymer that can absorb and retains a great extent of water, a property that is exploited in hydrogels making for various controlled drug delivery systems [1, 2]; a related fact is that the swelling behavior of PAA hydrogels was found to be highly medium pH dependent due to the presence of carboxylic groups [3]. Low molecular weight PAA has been valorized as an efficient salt-scaling inhibitor, chiefly in oilfield operations, via the ability of the carboxylic groups in sequestering metal ions [4]. A five-fold enhancement in the removal heavy metals was observed with PAA-bound tetrahydroxybenzene (THB-PAA) with respect to the virgin PAA [5]. Pyridylazo-2-naphthol-functionalized PAA was employed as a fluorescent nanofiber probe for the determination of Ni2+ [6]. Besides of being water soluble, PAA has a capacity to form polyelectrolyte complex nanoparticles with other polymers, such as poly(ethyleneimine) [7]. Ultrathin films for electrochromic devices were made via a layer-by-layer (LBL) strategy by reacting PAA with polyaniline in acid–base (COOH/NH2) mechanism [8]. By the virtue of its carboxylic groups, PAA served as template for making helical polyaniline-PAA microwires or rods [9]. The plausible decarboxylation of PAA via a homolytic pathway promoted its use as a redox unit carrier [10]. Apart from its hydrophilic character, PAA via its carboxylic groups was of a critical value in coating bleomycin-loaded magnetite nanoparticles (Fe3O4) to generate an antitumoral drug delivery system [11]. As for polycarboxylic acids [12], the other facet of PAA is its propensity towards potentiometric titration. The latter property is of a great merit, because the pH plays a pivotal role in several investigations, indicating the actual form of the involved molecule (ionized and neutral species). Moreover, the polymer effect emerges in various chemical studies as the results using polymers may be different from those using their molecular counterparts; for example, the potentiometric titration of PAA exhibits a profile different from those of acrylic acid and acetic acid.

In a continuation of a previous work [5, 13, 14], it is herein intended in the present paper to apply the chemistry of Minisci [15] at the macromolecular level, and henceforth to the chemical modification of poly(acrylic acid) (PAA) with hydroquinone (HQ); the latter one, being a polyhydroxylated benzene, is conceived as ligating species in the removal of heavy metallic ions via adsorption process. A synergism of COOH and HQ functionalities within the polymeric material in the removal of heavy metals is herein demonstrated; introducing HQ moieties onto PAA matrix would enhance the metal adsorption capacity of this polymer. Recently, we reported the results of adsorption of cadmium and led by COOH-functionalized mesoporous silica [16, 17]. Resins containing catechol and resorcinol were evaluated as sorbents for heavy metal ions [18, 19]. The present study includes the preparation of the HQ-PAA resin and an observation of the effects of reaction time and benzoquinone/monomer unit molar ratio on the degree of substitution, and of the effects of several experimental parameters, such as contact time, initial pH, electrolyte concentration, temperature, and initial resin mass, and copper concentration on the capacity exchange in the metal adsorption. The adsorption kinetics and isotherm models were used to describe the copper adsorption data. The thermodynamic factors of adsorption were determined.

Experiment

Materials and measurements

The chemicals were purchased from one of the following chemical companies: Aldrich (France), Fluka AG, Prolabo, and Merck. Acrylic acid was purified by vacuum distillation. Azobisisobutyronitrile (AIBN) was recrystallized from methanol. 1,4-Benzoquinone (BQ) was used as purchased. UV–vis spectra were recorded using a JASCO UV–Vis spectrophotometer using bidistilled water as a solvent. Infrared spectra were taken with a JASCO 4200 FT-IR. The polymer samples for the infrared analysis were in the form of cast films. The diagrams of X-ray diffraction (XRD) were recorded on a Bruker D8 diffractometer (K radiation, λ = 0.154 nm, 40 kV, 100 mA). The experiments were performed directly on the polymer samples which were finely cut in the range of 2θ equal to 5°–90° at the scanning speed of 0.02°/s. The morphology of polymeric samples was observed using a scanning electron microscope (SEM) JEOL 6700 operating at an acceleration voltage of 200 kV. 13C NMR spectra were recorded using a spectrometer Bruker 400 MHz using CD3OD as the solvent. Molecular weights were measured by size exclusion chromatography (SEC) [gel permeation chromatography (GPC)] on a set DIONEX, series Ultimate 3000 (degasser, pump, autosampler), 4 Shodex OH-pack 30 cm in series (803HQ, 804HQ, 806HQ, and 807HQ), a differential refractometer OPTILABrEX (Wyatt Tech), and a light-scattering detector DAWN multiangle HELEOSII (Wyatt Tech). The eluent was an aqueous solution consisting of water and 0.1 M NaNO3 at a rate of 0.5 mL min−1. The injected volume was 0.1 mL from a concentration of 2–3 g mL−1 of each polymer solution. Differential scanning calorimetry (DSC) thermograms were recorded on Netzsch DSC 204 F1 at a heating rate of 5 °C min−1 under nitrogen atmosphere. Thermogravimetric analysis (TGA) was conducted on NETZSCH TG 209 F3 apparatus in the temperature interval of 20–600 °C, at a heating rate of 10 °C min−1 under nitrogen atmosphere (2 mL min−1). The used weights of polymer samples for the measurement were in the range of 5.5–5.7 mg. Metallic adsorption by resins was estimated by flame atomic absorption spectrometer (FAAS) of type Buck Scientifics 210VGP fitted with a deuterium lamp, a corresponding hollow cathode lamp, and a 10-cm long burner; the flame was provided by air-acetylene gas.

Modification reactions of poly(acrylic acid) PAA (Eq. 1)

In a 500 mL, one-neck round bottom flask equipped with a condenser 0.5 g (6.94 × 10−3 mol) of PAA was dissolved in 50 mL of distilled water, followed by the addition of 50 mL of an ethanolic solution containing 0.93 g (8.6 × 10−3 mol) of BQ. To this homogeneous mixture, 0.039 g (2.3 × 10−4 mol) of silver nitrate was added, followed by the addition of 150 mL of distilled water. The system was heated to reach 75–80 °C under magnetic stirring. Once this temperature was reached, a saturated solution of potassium persulfate (1.88 g, 6.95 × 10−3 mol) was added dropwise. The reaction was stopped, and the content was cooled to room temperature. The modified PAA was isolated by the following sequential steps: (1) the evaporation of ethanol followed by the addition of 50 mL of distilled water; (2) the mixture was allowed to stand for 1 day to assure the precipitation of the unreacted BQ (the precipitate was eliminated by filtration); and (3) the modified PAA was obtained by casting from the solution within 4 days. The cast film was then dried to constant weight in vacuo at 45 °C for 4 days.

The modified PAA is herein abbreviated as HQ-PAA

(1)

.

The degree of substitution S (%) was provided by the following equations (Eqs. 2, 3) [10, 20]:

$$S\% = \frac{{n_{2} }}{{n_{1} + n_{2} }}$$
(2)
$$n_{1} = \frac{{m - n_{2} M_{2} }}{{M_{1} }}$$
(3)

where n 2 is the number of moles of HQ (in the form of vinyl hydroquinone) present in the modified PAA, estimated from the UV calibration curve of HQ; n 1 is the number of moles of acrylic acid present in the HQ-PAA, provided by Eq. (3); m is the sample weight of the HQ-PAA used in the UV measurements; M 1 and M 2 are the molecular weights of acrylic acid and vinyl hydroquinone, respectively.

Metallic ion adsorption

Preparation of the resins (Eqs. 4 and 5)

Into a 100 mL round-bottomed flask fitted with a reflux condenser, 5 g of polymer were charged and dissolved in 100 mL of distilled water. An amount of 0.39 g (2.3 × 10−3 mol) of AgNO3 was added, and the mixture was heated while stirring at a temperature of 70–75 °C, then a 100 mL aqueous solution of 1.88 g (6.96 × 10−3 mol) of K2S2O8 was added dropwise to the mixture. The latter was allowed to stir at this temperature domain for almost 1 h, the time after which a precipitate was observed. The precipitate was isolated and washed with a plenty of water and dried to a constant weight in vacuo for 4 days at a temperature of 45 °C.

(4)
(5)

Metal ion adsorption/desorption

In a thermostated cell, 0.1 g of cross-linked resin was added to 50 mL of 50 mg L−1 Cu(NO3)2 aqueous solution, which was previously adjusted to a desired pH (with 0.1 N HCl and 0.1 N NaOH). The system was stirred with a magnetic bar at a temperature of 30 °C. Sampling of the supernatant liquid was realized at specified time intervals. The adsorption capacity Q t (in mg of copper per gram of resin) [21] was computed from Eq. 6:

$$Q_{t} = \frac{{(C_{0} - C_{\text{e}} ){\cdot}V}}{W}$$
(6)

where C 0, C e, V, and W are the initial copper concentration in the aqueous solution (mg L−1), the copper concentration at equilibrium (mg L−1), the volume of aqueous solution (L), and the weight of the resin (g), respectively. A triplicate adsorption analysis was undertaken, and the mean values were considered.

The desorption of metal ion from the dried loaded resin (0.2 g) was conducted using 0.1 N HNO3 (50 mL) at 30 °C for 3 h. The regenerated resin was subjected to two other adsorption–desorption cycles. The desorption percentage DP [22] was calculated according to Eq. 7:

$$\varvec{DP} = \frac{\text{Amount of metallic ion desorbed in the medium}}{\text{Amount of metallic ion present in the resin}} \times 100.$$
(7)

Results and discussion

Chemical modification of PAA

Poly(acrylic acid) was synthesized as previously reported [5] and subjected to chemical reaction with p-benzoquinone by the virtue of the oxidative decarboxylation conditions of Minisci [13], as traced in Eq. 1. The mechanism proposed for such a chemical modification is depicted in Scheme 1 [23]. The system K2S2O8/AgNO3 induced the formation of a polymeric radical PAA·. Being an electron-rich site and behaving as a nucleophilic alkyl, the latter radical readily reacted with a benzoquinone compound, acting as an electron-deficient species. Minisci advanced such a mechanistic explanation for the homolytic alkylation of pyridine and its derivatives [24, 25]. Nucleophilic alkyl radicals were reported to attack preferentially the carbon atom of the benzoquinone compound [24].

Scheme 1
scheme 1

Mechanism of the grafting of hydroquinone (HQ) on poly(acrylic acid)

Full size image

Modified PAA displayed similar traits to the pristine polymer in terms of solubility, that is, HQ-PAA was soluble in polar solvents, such as water, methanol, dioxane, THF, and DMF and defied solubility in non-polar organic solvents.

Analyses by UV–visible, FT-IR, XRD, SEM, and 13C NMR (Figs. 1, 2, 3, 4, 5) certified the success of chemical modifications of PAA with HQ. UV–visible spectrum of HQ-PAA (Fig. 1) reveals a new absorption band at λ max = 288 nm, a featuring band of a hydroquinone unit. In addition, the infrared analysis (Fig. 2) came up with new absorption bands for HQ-PAA compared with those of virgin PAA. As can be clearly seen, it seems that the incorporation of HQ units onto PAA matrix has helped resolving its FT-IR spectrum, hinting at a high and regular substitution. The characteristic bands of the HQ entity appeared clearly as: (1) bands at around 1500–1550 cm−1, assigned to the aromatic C=C stretching; (2) two sharper bands at 860–870 and 1200–1250 cm−1, attributable to the angular deformation of C=C and the stretching of C–O, respectively; (3) a band at 1400–1425 cm−1, assigned to planar C–H band of the benzene ring; and (4) a wide band at around 3400 cm−1, ascribed to O–H band.

Fig. 1
figure 1

UV–visible spectra: a PAA; b HQ-PAA ([BQ]/[AA] = 1.25). Solvent: bidistilled water

Full size image
Fig. 2
figure 2

FT-IR spectra: a PAA; b HQ-PAA ([BQ]/[AA] = 1.25)

Full size image
Fig. 3
figure 3

13C NMR spectra of PAA and HQ-PAA. CD3OD as a solvent

Full size image
Fig. 4
figure 4

XRD spectra of PAA and HQ-PAA ([BQ]/[AA] = 1.25)

Full size image
Fig. 5
figure 5

SEM images: a PAA, ×230; b HQ-PAA, ×230; c HQ-PAA, ×1500

Full size image

The analysis by 13C NMR spectroscopy further supported the success of the chemical modification of poly(acrylic acid) (Fig. 3). Indeed, besides the chemical shifts of different carbons of PAA appearing at around of 36–42, 44–54, and 178–180 ppm, those of the HQ units are 112–116, 130, and 150 ppm.

Figure 4 illustrates the XRD patterns of PAA and HQ-PAA, indicating amorphous structures for both polymers as reported by other workers [26]. In the diffractogram of HQ-PAA, new peaks at 2θ equal to 27, 32, and 45° appeared, suggesting that a degree of crystallinity was ensued. Our results are consistent with the reported ones [27], confirming the grafting of HQ in the structure of the modified PAA.

SEM images of PAA and HQ-PAA are depicted in Fig. 5. The texture of PAA (Fig. 5a) is relatively flat and smooth, compared to that of HQ-PAA (Fig. 5b, c) which shows roughened surfaces filled with needles. This change proved the grafting of HQ within the PAA chains.

Surprisingly, the average molecular weight (Fig. 6; Table 1) of HQ-PAA was nearly unchanged or slightly lower than that of the starting PAA. Indeed, M w of the modified HQ x-PAA (x, a degree of substitution equal to 32 %) was estimated as 2.38 × 106 g mole−1 against 2.54 × 106 g mole−1 for PAA; that is, the mass of the modified polymer was about 1.06 times lower than the initial molecular weight of PAA, suggesting a degradation occurrence in the course of the modification reaction. This phenomenon was also observed earlier upon the reaction of PAA and PAAm (polyacrylamide) with 2,5-dihydroxybenzoquinone (DHBQ) under identical conditions [5, 13]. Neira et al. [28] reported the degradation of PAA when grafted onto cellulose using Fenton’s reagent (Fe(II)/H2O2), a redox system.

Fig. 6
figure 6

GPC elution profiles for PAA and HQ-PAA ([BQ]/[AA]) = 1.25)

Full size image
Table 1 Molecular weights of PAA, HQ-PAA, and their T g
Full size table

The glass transition temperature for the modified PAA (T g = 135.8 °C) was lower than that for the pristine PAA (151.26 °C), as noticed in Fig. 7 (Table 1). The lower T g measured for HQ-PAA can be due to the relatively low molecular weight obtained. The DSC thermogram of HQ-PAA revealed another thermal transition at 212.8 °C which could be taken as the melting point T m of the modified PAA. The small area at this endothermal peak would hint at the degree of crystallinity ensued during such functionalization as discussed above.

Fig. 7
figure 7

DSC thermograms of PAA and HQ-PAA ([BQ]/[AA]) = 1.25. Scanning rate: 5 °C/min; N2 atmosphere

Full size image

Thermogravimetric analyses of PAA and HQ-PAA are presented in Fig. 8. A two-step degradation for PAA can be noticed; an intense degradation of PAA (about 71 % weight loss) occurred between 250 and 300 °C, followed by a weaker one between 360 and 460 °C (about 25 % weight loss). However, a three-stage degradation of HQ-PAA was detected; the first stage (10 %) started at 100 °C and continued up to 160 °C, the second (35 % weight loss) between 200 and 280 °C, and the third (31 % weight loss) began at 360 °C and continued to 440 °C. Three maximum loss rates manifested at 230, 403, and 131 °C; the first loss rate may be attributed to the decarboxylation or anhydride formation, and the second and the third loss rates may be assigned to the degradation of HQ moieties. The results obtained by TGA for HQ-PAA and PAA were consistent with those revealed by DSC measurements.

Fig. 8
figure 8

TGA thermograms of PAA and HQ-PAA ([BQ]/[AA]) = 1.25. Heating rate: 10 °C/min; N2 atmosphere

Full size image

Effect of [BQ]/[monomer unit] ratio

The effect of [modifier]/[monomer unit] molar ratio, [BQ]/[AA], on substitution extent of the modification of PAA was assessed at 70–80 °C and for two reaction times, 3 and 6 h. Such an effect is illustrated in Fig. 9. The degree of modification was estimated by UV absorbance of the HQ-PAA using a calibration curve that was established by measuring the absorbance of dilute hydroquinone solutions at λ max = 288 nm; as can be noticed in the figure, the variations of modification degree increased with increasing [BQ]/[AA] and high substitution extents (29–32 %) were gained with a ratio higher than 1. In addition, prolonging the reaction time from 3 to 6 h in both the cases did not improve significantly the substitution extent.

Fig. 9
figure 9

Plot of the variation of substitution degree of PAA, S (%), as a function of [BQ]/[AA] molar ratio. Temperature = 70–80 °C

Full size image

Effect of reaction time

The impact of reaction time on the modification extent of PAA for a [BQ]/[AA] ratio of 1.25 is presented in Fig. 10. The substitution extent increased with increasing reaction time, and optimal substitutions were reached at a time of 5 h and leveled off at 32 % beyond this time.

Fig. 10
figure 10

Variation of the substitution degree of PAA, S (%), as a function of time. T = 70–80 °C

Full size image

Metallic adsorption–desorption study

PAA and HQ-PAA adsorbents were set by cross-linking their corresponding soluble polymers applying the oxidative decarboxylation conditions of Minisci as traced in Eqs. 4 and 5. The HQ-PAA used in the adsorption of the metallic ions was the one modified with a molar ratio [BQ]/[AA] of 1.25 that corresponds to 32 % degree of substitution. The effects of pH, time, ionic strength, temperature, adsorbent mass, and initial concentration of copper, on the Cu(II) adsorption capacities of the PAA and HQ-PAA resins, were assessed, and the results are illustrated in Figs. 11, 12, 13, 14, and 15, 16 and compiled in Tables 2, 3, 4, and 5. A tentative structure of the HQ-PAA adsorbent/Cu2+ complex at pH of 5.4 is pictured in Fig. 11.

Fig. 11
figure 11

Proposed structure of the copper (II)/HQ-PAA complex at pH 5.4

Full size image
Fig. 12
figure 12

Variations of adsorption capacity Q t of PAA resin for copper (II) as a function of time and pH. V = 50 mL, C 0 = 50 mg/L, W = 0.1 g, [NaNO3] = 0, and T = 30 °C

Full size image
Fig. 13
figure 13

Variations of adsorption capacity Q t of HQ-PAA resin ([BQ]/[AA]) = 1.25 for copper (II) as a function of time and pH. V = 50 mL, C 0 = 50 mg/L, W = 0.1 g, [NaNO3] = 0, and T = 30 °C

Full size image
Fig. 14
figure 14

Variations of adsorption capacity Q t of HQ-PAA resin ([BQ]/[AA]) = 1.25 for copper (II) as a function of time and [NaNO3]. V = 50 mL, C 0 = 50 mg/L, W = 0.1 g, pH 5.4, and T = 30 °C

Full size image
Fig. 15
figure 15

Variations of adsorption capacity Q t of HQ-PAA ([BQ]/[AA]) = 1.25 and PAA resins for copper (II) as a function of time and temperature. V = 50 mL, C 0 = 50 mg/L, W = 0.1 g, [NaNO3] = 0, T = 30 °C, and pH 5.4

Full size image
Fig. 16
figure 16

Variations of adsorption capacity Q t of HQ-PAA resin ([BQ]/[AA]) = 1.25 for copper (II) as a function of adsorbent mass. V = 50 ml, C 0 = 50 mg/L, pH 5.4, and T = 30 °C

Full size image
Table 2 Sorption of copper ion at different [NaNO3]
Full size table
Table 3 Sorption of copper ion at different temperatures
Full size table
Table 4 Sorption of copper ion at different sorbent masses
Full size table
Table 5 Sorption of copper ion at different Cu2+ concentrations
Full size table

It can be seen that the adsorption capacities of the PAA and HQ-PAA resins with respect to copper (II) increased with increasing pH and time, as shown in Figs. 12 and 13. They remained constant beyond a time of about 3 h. At pH values lower than 3, the sorption rate increased by raising the pH, but at higher values, the rate was independent on pH. The impact of this parameter on the removal of metallic ions from solutions is clear, because it affects the surface charge of sorbent. A higher pH would favor the tendency of negative charge formation on the sorbent, leading to an increase in the electrostatic attraction between the positively charged metallic ions and the negatively charged sorbent, thus a greater metallic sorption. The capacity exchanges were found to be 3.58 and 17.17 mg g−1, respectively, at an optimal pH of 5.4 and a time of 3 h. Hence, HQ-PAA resin was nearly five times more effective than PAA resin in terms of copper (II) adsorption. This finding can be attributed to the grafted HQ moieties on the PAA matrix. The experiment was limited to the indicated pH values, because copper hydroxide would precipitate at pH values higher than 5.4, and would interfere with the actual results. On the other hand, the removal efficiency of metallic ion was found unexpectedly unaffected by increasing the ion strength of the solution, as shown in Fig. 14 and Table 2. In fact, the addition of NaNO3, as an electrolyte, at varying concentrations neither improved nor declined the sorption property. The present results are in good agreement with those reported [29, 30].

The effect of temperature on the adsorption capacities of the resins is shown in Fig. 15 and Table 3, demonstrating that the higher the temperature, the greater the adsorption capacity; at optimal pH (5.4) and time (3 h), the exchange capacities were 5.5 and 20.25 mg g−1 at a temperature of 45 °C for PAA and HQ-PAA resins, respectively. In these conditions, the COOH groups in the adsorbing resins are in the form of carboxylate ions, because the working pH (5.4) is higher than pKa of PAA (4.7), and the phenolic OH groups in the HQ-PAA resin are not ionized due to their higher pKa (~9.4); the carboxylate ions would complex metallic ions more favorably than the catechol hydroxyl groups.

Figures 16 and 17 picture the effects of the adsorbing resin mass and the initial copper concentration on the sorption of copper ion Cu(II), and the results are compiled in Tables 4 and 5. As can be noticed in Table 4, about 94 % enhancement of the exchange capacity of PAA resin was observed at optimal conditions by increasing the mass of the sorbent by twofold, but only 22 % enhancement for HQ-PAA resin was found. Raising the initial concentration of copper nitrate improved the metal ion uptakes of the resins (Table 5) with nearly the same extents as with the variation of sorbent mass. These results are in tune with the published reports [31, 32].

Fig. 17
figure 17

Variations of adsorption capacity Q t of HQ-PAA resin ([BQ]/[AA]) = 1.25 for copper (II) as a function of Cu(II) concentration. V = 50 mL, pH 5.4, T = 30 °C, and W = 0.1 g

Full size image

The adsorption of lead (II) and cadmium (II) by the resins was conducted and compared with that of copper (II) under the same optimal conditions. The adsorption capacities of the resins towards these metal ions are depicted in Fig. 18 and gathered in Table 6. Maximum adsorptions of Pb(II) and Cd(II) by PAA resin were 7.04 and 2.88 mg g−1, and were 20.84 and 12.69 mg g−1 by HQ-PAA resin. Under the actual experimental conditions, the exchange capacity is in this order: Pb(II) > Cu(II) > Cd(II). The adsorption capacity of the adsorbent is more for lead when compared to other metals, because the stability constants of complexes of lead with carboxylates and phenols are greater than those for other metals [33, 34]. The cross-linked poly(acrylic acid-co-Kryptofix-22-DM) [36], the tetrahydroxybenzene-functionalized PAA [5], and the tetrahydroxybenzene-functionalized PAAm [14] showed the same order of sorption capacity. Overall, our results are within with those reported in the literature (Table 7).

Fig. 18
figure 18

Variations of adsorption capacities Q t of HQ-PAA ([BQ]/[AA]) = 1.25 and PAA resins for copper (II), lead (II), and cadmium (II) as a function of time. V = 50 mL, C 0 = 50 mg/L, W = 0.1 g, [NaNO3] = 0, T = 30 °C, and pH 5.4

Full size image
Table 6 Sorption of Pb2+, Cu2+, and Cd2+
Full size table
Table 7 Adsorption capacity (mg g−1) of some polyhydroxybenzene-functionalized resins towards Cu2+, Cd2+, and Pb2+
Full size table

Adsorption kinetics

Pseudo-first-order model (Lagergren equation)

The pseudo-first-order kinetic is expressed, as shown in Eq. 8 [37]:

$$\log (Q_{\text{e}} - Q_{t} ) = \log Q_{\text{e}} - \frac{{k_{1} t}}{2.303}$$
(8)

where

k 1 is the pseudo-first-order rate constant (min−1),

Q e is the adsorption capacity of the adsorbent (mg g−1),

Q t is the amount of Cu ion adsorbed at time t (mg g−1).

where k 1 is computed from the slope of the plot of ln(Q e − Q t ) versus t (Fig. 19).

Fig. 19
figure 19

Pseudo-first-order model fit for the adsorption of Cu2+ at pH 5.4: a PAA; b HQ-PAA

Full size image
Pseudo-second-order model (Ho-McKay equation)

The pseudo-second-order kinetic is expressed as shown in Eq. 9 [38].

$${\raise0.7ex\hbox{$t$} \!\mathord{\left/ {\vphantom {t {Q_{\text{t}} }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${Q_{\text{t}} }$}} = {\raise0.7ex\hbox{$t$} \!\mathord{\left/ {\vphantom {t {Q_{\text{e}} }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${Q_{\text{e}} }$}} + {\raise0.7ex\hbox{$1$} \!\mathord{\left/ {\vphantom {1 {k_{2} Q_{\text{e}}^{2} }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${k_{2} Q_{\text{e}}^{2} }$}}$$
(9)

where k 2 is the pseudo-second-order rate constant (g mg−1 min−1) and computed from the slope of the plot of t/Q t versus t (Fig. 20).

Fig. 20
figure 20

Pseudo-second-order model fit for the adsorption of Cu2+ at pH 5.4: a PAA; b HQ-PAA

Full size image

The results of the applied models to the copper (II) adsorption by PAA and HQ-PAA are illustrated in Figs. 19 and 20 and compiled in Table 8. As can be noticed in Table 8, the experimental Q e values are so distant from the calculated ones and the correlation coefficients are low, suggesting that this adsorption does not obey a pseudo-first kinetic model; the graphs in Fig. 19 are not straight lines, precluding the validity of Lagergren equation. However, the straight lines of the plot t/Q t versus time (Fig. 20) are indicative of the validity of Ho-McKay equation; indeed, the experimental Q e values and the calculated ones are so close and the correlation coefficients are high as shown in Table 8, suggesting that the adsorption fits well a pseudo-second kinetic model.

Table 8 Results of pseudo-first- and second-order models of Cu2+ adsorption
Full size table

Adsorption isotherms

Three adsorption isotherms models were examined for our data on the copper adsorption by PAA and HQ-PAA: Freundlich, Langumir, and Temkin.

Freundlich isotherm

The linear form of the Freundlich equation is given in Eq. 10 [39]. For Freundlich isotherm, the ratio of sorbate adsorbed onto a given mass of adsorbent to the sorbate concentration in solution is not constant:

$$\log Q_{\text{e}} = \log K_{\text{F}} + (1/n)\log C_{\text{e}}$$
(10)

where

Q e is the amount of solute adsorbed per-unit weight of adsorbent (mg g−1),

K F is the Freundlich constant (mg g−1),

C e is the equilibrium concentration of the solute in the bulk solution (mg L−1),

1/n is the heterogeneity factor indicating the adsorption intensity.

Langumir isotherm

The linear form of Langmuir isotherm equation is given in Eq. 11 [40]. For the Langmuir isotherm, a monomolecular layer is supposedly formed, as the sorbate molecules are adsorbed without any interactions between them:

$${\raise0.7ex\hbox{${C_{\text{e}} }$} \!\mathord{\left/ {\vphantom {{C_{\text{e}} } {Q_{\text{e}} }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${Q_{\text{e}} }$}} = \left( {{\raise0.7ex\hbox{$1$} \!\mathord{\left/ {\vphantom {1 {Q_{\hbox{max} } K_{\text{L}} }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${Q_{\hbox{max} } K_{\text{L}} }$}}} \right) + \left( {{\raise0.7ex\hbox{${C_{\text{e}} }$} \!\mathord{\left/ {\vphantom {{C_{\text{e}} } {Q_{\hbox{max} } }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${Q_{\hbox{max} } }$}}} \right)$$
(11)

where

Q max is the maximum adsorption capacity of the adsorbent (mg g−1),

K L is the Langmuir adsorption constant (L mg−1).

Temkin isotherm

The linear form of the Temkin model equation is given in Eq. 12 [41]. For the Temkin model theory, the effect of the interactions between sorbate molecules on adsorption isotherms is taken into account, suggesting that the heat of adsorption of all the molecules in the layer would decrease linearly:

$$Q_{\text{e}} = B\ln A + B \ln C_{\text{e}}$$
(12)

where

B = RT/b T, b T is the Temkin constant related to heat of sorption (J mol−1), R the gas constant (8.314 J/mol K), and T is the absolute temperature,

A is the Temkin isotherm constant (L g−1).

Data of the study of adsorption isotherms are illustrated in Figs. 21, 22, and 23, and Table 9. Based on the correlation coefficient R 2, none of the examined isotherm models would fit the adsorption data in the case of PAA; indeed, R 2 is poorly low and is in the range of 0.40–0.76. However, in the case of HQ-PAA, the three models are more favorable; yet, the Langmuir model (R 2 = 0.993) fits better than Temkin (R 2 = 0.918) and Freundlich (R 2 = 0.816). The fact that the present data obey better to Langmuir model can be assigned to the homogenous distribution of active sites on the HQ-PAA surface, as the Langmuir equation assumes a homogenous surface, where all the sites have equal adsorption energies. The Langmuir constant K L is 0.181 L mg−1, a value lower than 1, indicating favorable adsorption. The maximum adsorption capacity of the HQ-PAA adsorbent, Q max, was measured to be 22.37 mg g−1. The Freundlich treatment of our data afforded 1/n values greater than 1, suggesting that the Cu2+ adsorption onto HQ-PAA adsorbent was unfavorable under the studied experimental conditions.

Fig. 21
figure 21

Plots of Freundlich model for Cu2+ adsorption: a PAA; b HQ-PAA

Full size image
Fig. 22
figure 22

Plots of Langmuir model for Cu2+ adsorption: a PAA; b HQ-PAA

Full size image
Fig. 23
figure 23

Plots of Temkin model for Cu2+ adsorption: a PAA; b HQ-PAA

Full size image
Table 9 Parameters of adsorption isotherm models for Cu2+/PAA and Cu2+/HQ-PAA systems
Full size table

Thermodynamic adsorption parameters

To acquire an insight into the adsorption process of the Cu2+ ions and its distribution between the adsorbent and the aqueous medium, thermodynamic factors of the whole process are eagerly sought for; the Gibbs free energy of adsorption (∆G0), the entropy of adsorption (∆S0), and the enthalpy of adsorption (∆H0) can be computed from the following equations (Eqs. 13–15). ∆H0 and ∆S0 were taken as slopes and intercepts of the plot of ln K d versus 1/T (Fig. 24), respectively. The results are gathered in Table 10. The positive value of enthalpy ∆H0 indicates that the adsorption is an endothermic process, and the positive value of entropy ∆S0 is attributed to an occurrence of a less ordered molecular arrangement of both carboxylate ion and hydroquinone unit [42]. Negative values of Gibbs free energy (∆G0) indicate a favorable and spontaneous adsorption. Worthy of note is that, at 20 °C, no adsorption took place as the value of ∆G0 is positive. At a working temperature, that is 30 °C, adsorption occurred to a certain extent as ∆G0 is −0.327 kJ mol−1; however, a more favorable and spontaneous adsorption could happen at 40 °C (∆G0 = −2.001 kJ mol−1):

$$\Delta {\text{G}}^{0} = - RT\ln K_{\text{d}}$$
(13)
$$\ln K_{\text{d}} = \frac{{\Delta {\text{S}}^{0} }}{R} - \frac{{\Delta {\text{H}}^{0} }}{RT}$$
(14)
$$K_{\text{d}} = \frac{{C_{0} - C_{e} }}{{C_{e} }}. \frac{V}{w}$$
(15)

where

Fig. 24
figure 24

Plots of lnK d versus 1/T. a PAA; b HQ-PAA

Full size image
Table 10 Thermodynamic parameters of the adsorption de Cu2+
Full size table

R is the ideal gas constant (8.314 J mol−1 K−1),

T is the absolute temperature (K),

K d is the equilibrium constant,

C e is the equilibrium metal ion concentration (mg L−1).

Desorption performance

In the sake of reusability of the resin in adsorption of metals, the adsorption–desorption of metal ions was carried out. The desorption of metal-adsorbed resins was accomplished by treatment with 0.1 N nitric acid at 30 °C. The results of adsorption–desorption of ions by the resins are gathered in Table 11. The results revealed that the capacity exchanges of the resins were retained even after three adsorption–desorption cycles for copper (II), lead (II), and cadmium (II) ions, hinting at the feasibility of reuse of the resins in the removal process. The desorption extents of these ions from PAA and HQ-PAA resins were substantially high and exceeding 98 %.

Table 11 Adsorption–desorption results
Full size table

Conclusions

The application of oxidative decarboxylation of Minisci to the chemical modification of poly(acrylic acid) was successful. As expected, metallic adsorption of hydroquinone-grafted poly(acrylic acid) was greater than that of the virgin poly(acrylic acid). In addition, to achieve higher metal ion uptakes by poly(acrylic acid) and its modified resin, one needs to work under the following conditions: pH 5.4, time 180 min, and temperature 45 °C. Metallic adsorption by hydroquinone-modified poly(acrylic acid) resin occurred at temperatures of only higher than 30 °C. Finally, the adsorption–desorption results suggest the utilization of the hydroquinone-modified poly(acrylic acid) as a cation exchange resin.