Introduction

The presence of organic compounds in the environment is of increasing concern due to their eco-toxicological effects. Some of these organic compounds are pharmaceutical agents with therapeutic importance, but when they get into the environment, especially water bodies, they become pollutant either directly or indirectly depending on their metabolites. Excessive use of pharmaceutical agents can create serious problems to the environment. The human or animal body may not be able to completely breakdown some of the dose when taken, and as a result such dose will have to be excreted as active compound [1]. On the other hand, a few of these pharmaceutical agents can also serve as reagents for biochemical reactions. When used as reagent, removal of these pharmaceutical agents after reaction is a concern. They are sometimes in industrial effluents from pharmaceutical industries, and there are cases where they are in laboratory wastes when used as reagents during chemical or biochemical reactions. This is a major problem in developing countries like Brazil and Nigeria. Presently in Brazil and Nigeria, there is need to develop an effective means of treating effluents from hospitals, pharmaceutical industries and laboratories (most especially academic institution based laboratories).

4-Aminoantipyrine is an example of pharmaceutical agent that serves as drug as well as biochemical reagent. Apart from being able to act as an analgesic and anti-inflammatory agent, 4-aminoantipyrine is used in the assay for aspartate aminotransferase or glutamate oxalacetate transaminase. Although it poses the risk of agranulocytosis, 4-aminoantipyrine can measure extracellular water. Studies have shown its use as a redox indicator during the determination of free and captured glucose oxidase [2, 3]. Studies have revealed the significance of its derivatives as biological and pharmacological agents [4, 5]. Currently, 4-aminoantipyrine is receiving large attention because of the important role it plays as potential pharmaceutical and biochemical reagent. Due to its continuous and excessive use, its presence has been detected in water at low concentration, which has created a serious concern. Developing an effective process for purifying water contaminated with low concentration of 4-aminoantipyrine is a challenge in developing nations, and this is the core focus of this present study.

Use of adsorption as a water purification process is economical and capable of removing organic pollutants from water system. Adsorption has played important role in the removal of low concentrations of organic pollutants from wastewater, aqueous solution, potable water and process effluents [6]. Although several materials such as activated carbon, zeolite and biomass are used as adsorbent to achieve this, but some of them have shortcomings such as inefficient removal, high cost, inefficient desorption and regeneration. There is need to develop new adsorbent that will circumvent these shortcomings. There may also be need for search for cheap, readily available and environmentally friendly materials that may be directly used or serve as base material for this purpose. Use of clay materials may meet this need as they can serve as viable adsorbent for the adsorption of organic compounds.

Vermiculite is a type of clay that may be suitable for this use [7]. It is a hydrous phyllosilicate mineral and trioctahedral with negative charges resulting from the substitution of Si4+ by trivalent cations in tetrahedral positions [8,9,10]. Depending on the water interlamellar layers and interlayer cations, vermiculites have a structural unit thickness of about 1.4 nm with the water interlamellar layers susceptible to hydration and dehydration processes [11]. The interlayer cations—Mg2+ and minor amounts of Ca2+, Na+, and K+ play a key role in the hydration properties, influenced by cation radius and charge. The cation exchange capacity (CEC) varies between the 120 and 200 cmol+/kg for air-dried vermiculites or 140 and 240 cmol+/kg for dehydrated vermiculites [10, 12]. Figure 1 presents the structural model of vermiculite from Brazil [10]. Vermiculite is abundantly present in Brazil, and there is need to find application for this naturally abundant resource in this part of the world. Use of vermiculite as a means of purifying 4-aminoantipyrine contaminated water will be an excellent way of finding application for this abundant resource in Brazil and other developing countries like Nigeria.

Fig. 1
figure 1

Model of the structure of vermiculite from Brazil (Valášková and Martynková [10])

Full size image

Vermiculite expands significantly when heated which indicates that the properties of vermiculite may improve via intercalation or through other chemical reactions. Previous works have shown that it can be expanded to produce highly porous materials for technological and environmental applications [13,14,15,16,17]. Modifying the surface properties of vermiculite is a key way to improve its capacity for specific applications such as in water purification. As an adsorbent, it may be modified by various means with the objective of increasing the adsorption capacity. This is achievable by enhancing the bond strength between the adsorbate and vermiculite (the adsorbent). Previous studies have shown that facile dispersion of cellulose nanocrystals (CNCs) on the surface and internal layers of expanded vermiculite can improve its properties and capacity for specific applications [17, 18]. Based on the ability of vermiculite to expand, the present study aims at improving the wastewater purifying capacity of vermiculite by modification via facile dispersion and intercalation of the expanded vermiculite with CNCs followed by surface functionalization with NTA.

Experimental

Materials

As previously reported [16], vermiculite was supplied by Vermiculita Ind. Com. (Brazil) with granulometry 0.3–0.6 cm and approximate composition (Al0.30Ti0.04Fe0.63Mg2.00) (Si3.21Al0.79)(O10(OH)2Mg0.13Na0.02K0.10(H2O)n). NTA and all other chemicals used in this study were purchased from Sigma-Aldrich (Belo Horizonte, Brazil) while 4-aminoantipyrine was purchased from Sigma-Adrich (Germany). Cellulose was obtained from eucalyptus kraft wood pulp with high alpha-cellulose content (96–98%).

Preparation of CNCs

The CNCs were prepared from cellulose via sulfuric acid hydrolysis as described by de Mesquita et al. [19]. Briefly, cellulose was added to 64 wt% sulfuric acid under vigorous mechanical stirring using a Fisatom mechanical stirrer for 25 min. After hydrolysis, the dispersion was diluted tenfold, and the suspensions washed using two-repeated centrifuge cycling at 7000 rpm for 10 min. This was dialyzed with deionized water to remove any non-reactive sulfate groups, salts and sugar until the solution reached pH 7.0. The purified suspension was ultrasonicated using a Cole Parmer ultrasonic processor (model CV 334, 20 kHz) for 30 min. Powdery CNCs were finally obtained by freeze-drying.

Preparation of VCM

VCL (38 g) of 3 m mesh size was dipped in CNCs (0.88%) for 1 h. After the dipping, the VCL–CNCs obtained were dried at 65 °C for 3 h in an oven. This step of dipping VCL in CNCs was repeated four times (4-cycles). NTA was imprinted on VCL–CNCs by simple reaction. To achieve this, 10 g of NTA was placed in a two-necked round bottom flask containing 100 mL deionized water. The temperature was kept at 60 °C and stirred for 1 h. VCL–CNCs was accurately weighed and poured into the flask, while the temperature was gently raised to 70 °C. Sodium methoxide was added, while refluxing for 18 h. The final product was filtered, washed severally with deionized water and oven-dried at 50 °C to obtain the VCM.

Characterization

VCL and VCM were blended with KBr, pressed into pellets and analyzed for their functional groups composition using FTIR (FTIR, Perkin Elmer, spectrum RXI 83303, MA, USA) in the range of 400–4500 cm−1. Thermogravimetric analysis of VCL and VCM was carried out using DTA-TG equipment (C30574600245, Shimadzu, Tokyo, Japan). The X-ray diffraction pattern of VCL and VCM was recorded from 10° to 80° (2θ) with a scanning speed of 2.00°/min using X-ray diffractometer (XRD-7000X-Ray diffractometer, Shimadzu) with filtered Cu Kα radiation operated at 40 kV and 40 mA. The surface area was determined by nitrogen adsorption using the BET method in a Quantachrome Autosorb IQ2 instrument while surface morphology was obtained using SEM–EDS (FEI quanta 200 3D, dual beam FEG, model EDAX EDS brand, Bruker) with powdered samples being coated with gold using the sputtering technique in order to increase electrical conductivity and the quality of the micrographs. Particle size distribution and zeta potential were recorded using a zeta potential analyzer (DT1200, Dispersion technology) at 25 °C while observing general calculation model for irregular particles. Measurements of the zeta potential and particle size distribution were read using Dispersion technology-AcoustoPhor Zeta size 1201 software (version 5.6.16).

Adsorption experiment

Batch adsorption experiments were performed at room temperature (298 K) on a rotary shaker using 0.1 g of adsorbent (VCL or VCM) in 250 mL Erlenmeyer’s flasks at different initial concentrations of 4-aminoantipyrine (100 mL) ranging from 1 to 5 mg L−1. All the experiments were performed at a shaking speed of 200 rpm for 3 h. The experimental equilibrium time of 3 h was established after several experimental trials at 298 K and 200 rpm. The concentration of 4-aminoantipyrine was monitored by spectrophotometric analysis using a Schimadzu UV–Vis spectrophotometer (model: UV-1650 PC, Schimadzu corporation, Kyoto, Japan) at a wavelength of 244 nm. The amount of 4-aminoantipyrine adsorbed by VCL and VCM was calculated using equation:

$$ q_{\text{e}} = \frac{{\left( {C_{\text{o}} - C_{\text{e}} } \right)V}}{M} $$
(1)

where qe is the amount adsorbed in mg g−1, \( C_{\text{o}} \) and \( C_{\text{e}} \) are initial and final concentrations (mg L−1) of 4-aminoantipyrine in solution, respectively; V and M are volumes (L) of 4-aminoantipyrine solution and weight (g) of adsorbent (VCL and VCM) used.

Desorption studies

The desorption study was carried out in the presence of ethanol, distilled water, benzene, 0.5 M HCl, 0.5 M NaOH and acetone. VCL and VCM were loaded with 4-aminoantipyrine by contacting 0.1 g of VCL or VCM with 4-aminoantipyrine (1 mg L−1) in 250 mL Erlenmeyer’s flasks at room temperature. This was performed at a shaking speed of 200 rpm for 2 h. The 4-aminoantipyrine loaded VCL and VCM were dried at room temperature for 72 h. In order to determine the desorption capacity, 4-aminoantipyrine loaded VCL and VCM were separately poured into different 250-mL Erlenmeyer’s flasks containing 100 mL ethanol, distilled water, benzene, 0.5 M HCl, 0.5 M NaOH and acetone. These were shaken at room temperature and at a speed of 200 rpm for 2 h. The desorbed 4-aminoantipyrine was measured by spectrophotometric analysis using a Schimadzu UV–Vis spectrophotometer (model: UV-1650 PC, Schimadzu corporation, Kyoto, Japan) at a wavelength of 244 nm. Desorption of 4-aminoantipyrine from VCL and VCM was calculated as:

$$ {\text{Desorption}}\,\left( \% \right) = \frac{{q_{\text{e}}\, \left( {\text{desorption}} \right)}}{{q_{\text{e}}\, \left( {\text{adsorption}} \right)}} \times 100 $$
(2)

Quantum chemical parameters

All theoretical calculations were performed using the DFT electronic structure programs at B3LYP/6-31G level theory using Spartan 14.1 software. The molecular electronic structure of 4-aminoantipyrine was modeled, including the distribution of frontier molecular orbitals in order to establish the reactivity of 4-aminoantipyrine as well as its active sites for the adsorption process.

Results and discussion

Preparation of VCM and characterization

The preparation of VCM is described as shown in Fig. 2. Vermiculite is a phyllosilicate mineral, which is hydrous in nature. When heated, it expands quickly to produce a lightweight material. If the heating is rapid, vermiculite exfoliates as the interlayer water turns into steam, forcing the silicate layers apart from one another in an accordion like expansion (VCL in Fig. 2) with expansion ratio of 20–30 times its original thickness depend on the temperature and time of processing [20, 21]. This expansion of vermiculite on heating allows the dispersion of the CNCs into the space created from the expansion. CNCs were deposited on the surface of VCL. This facile dispersion and intercalation of VCL may be due to the capacity of CNCs to facially diffuse throughout VCL and the possibility of hydrogen bonding taking place between VCL and CNCs. The functionalization may be explained as being esterification reaction and facial diffusion. The reaction was catalyzed by the sodium methoxide. Since both VCL and CNCs contains hydroxyl (OH) functional group, the functionalization took place between the OH group and the carboxylic group of the NTA.

Fig. 2
figure 2

Schematic representation of the production of VCM

Full size image

Figure 3 reveals the functional groups present in VCL, CNCs and VCM. Bands were observed at 3564 and 3419 cm−1 in both VCL and VCM, which were attributed to the –OH functional group of the interlayer water molecules and the silanol groups. Presence of –OH functional group was also found in CNCs at 3406 cm−1. The band for the –OH functional group is broader in VCM than VCL which may be due to the contribution of –OH group from CNCs and NTA. All the samples exhibited a peak at 1640 cm−1 that was considered as being due to the –OH bending vibrational frequency of internally bonded water molecules present in the samples. Both VCL and VCM showed strong bands at 1012 cm−1, which were related to the Si–O stretching vibration. They both also exhibited weak bands at 756 and 667 cm−1 which were attributed to bending vibrations of Al–OH and Al–O, respectively [22]. The Si–O–Mg bending vibration was also seen in VCL and VCM at 451 cm−1 [23]. The band at 1058 cm−1 in CNCs and VCM was assigned to C–O–C pyranose ring stretching vibration, and this band is absent in VCL which suggests the presence of CNCs in VCM. The vibrational frequencies found at 1163 cm−1 and 1112 cm−1 in CNCs were considered as being due to the C–C ring breathing band and C–O–C glycosidic ether band from the polysaccharide component, respectively. These bands were found in VCM but not in VCL; the presence of these bands further suggests the presence of molecules of CNCs in VCM. Bands at 2921 cm−1 in VCM and CNCs were attributed to C–H stretching of CH2 group while bands at 1315 cm−1 was assigned to the bending vibration of C–O bonds in the polysaccharide rings [24]. The bands at 1411 and 1327 cm−1 found only in VCM were due to the carboxylic acid deprotonated strong symmetric stretching of COO− group emanating from the surface functionalization of VCM with NTA.

Fig. 3
figure 3

FTIR of VCL, CNC and VCM

Full size image

Figure 4 shows the XRD analysis of VCL, CNCs and VCM, which presented broad and weak reflections. The result indicated that VCL and VCM contains mainly vermiculite, small amounts of mica, illite and hydrobiotite. The pattern shows multiple reflections, suggesting mixture of different minerals as previously reported [25, 26]. The basal spacing (d001) of VCM is larger than that of VCL which suggested the dispersion of CNCs in the layers of VCL to form VCM. The diffractogram further revealed a prominent reflection at 22.60° (X) in CNCs which appeared at 24.20° (X) in VCM indicating the intercalation of the starting material with CNCs. The modification with CNCs and NTA decreased the intensity of basal reflections of VCL. This observation may be due to the partial transformation of the ordered structure of vermiculite clay into delaminated structure [27]. This may have facilitated the intercalation of bigger chemical species like CNCs. Smaller chemical species like NTA are capable of acting as chelating agents toward exchangeable cations in clays [28,29,30]. So, apart from NTA being involved in esterification reaction with the freely available hydroxyl groups on CNCs dispersed in the layers of VCL and also with the hydroxyl group on the surface of VCL, NTA is also capable of complexing the exchangeable cations present in VCL to form VCM.

Fig. 4
figure 4

XRD pattern of VCL, CNC and VCM (V: vermiculite, M: mica, I: illite, HB: hydrobiotite, X: CNC)

Full size image

The BET surface area of VCL was 4 m2 g−1 as previously reported [17] but after the modification, the surface area increased to 96 m2 g−1 in VCM as shown in Fig. 5a. The tremendous increase in surface area may be accounted for as being due to the delamination of structure and development of porosity in VCM that may have resulted in change of crystalline structure creating micropores and mesopores [31, 32]. The TGA curve of VCL, CNCs and VCM is presented in Fig. 5b. The first loss in mass was noticed at temperature below 100 °C in VCL, CNCs and VCM, which was attributed to loss of water molecules physisorbed on the surface of these materials. The second loss in mass was in the range 100–200 °C that may be due to loss of internally bonded water molecules in CNCs and loss of more free water molecules in the particle spaces of VCL and VCM. Loss observed in the range 205–340 °C in CNCs may be attributed to degradation leading to 1,4 and 1,6 anhydroglucopyranoside and depolymerization at 1,4 glycosidic bond [32]. Similar mass loss was only found in VCM at range 225–350 °C. Loss in mass at 450–800 °C in VCL may be attributed to dehydroxylation [33] while in the case of VCM it may be accounted for as being due to dehydroxylation as well as loss of char emanating from the modification with CNCs and NTA.

Fig. 5
figure 5

a BET of VCM, b TG curves of VCL, CNC and VCM, c PSD of VCL, d PSD of VCM, e PSD of CNC

Full size image

The particle size distribution is monomodal for the entire sample with mean size of 0.0359 µm in VCL, which increased to 10.1600 µm in VCM after the modification as shown in Fig. 5c–e. The zeta potential and SEM images are presented in Fig. 6. The zeta potential increased as the pH value increased in both VCL and VCM. These values help in quantifying the charges, degree of stability and possible electrostatic interaction at these surfaces of VCL and VCM [34, 35]. This distribution showed that VCL and VCM are stable in liquid medium with respect to solid–liquid interactions. The result presents a high zeta potential for the samples over the studied pH range. The high values correspond with previously reported high layer charges characterization for vermiculites [27], although this value reduced after the modification which may be due to the functionalization with NTA which may have complexed some of the cations at the surface of VCL. The SEM revealed both VCL and VCM to have flaky look with irregular shapes. These flakes produced some amounts of porosity in VCL before the introduction of CNCs and NTA. The dispersion of CNCs on the surface and internal layers of VCL was possible due to the nano-size dimension of CNCs and the hydrogen bonding between CNCs and the surface of VCL. The modification had influenced the densification and porosity in that few pores are on VCL as seen in the micrograph whereas more of the pores are in VCM.

Fig. 6
figure 6

Zeta potential and SEM images of VCL and VCM

Full size image

Adsorption process

Effect of contact time and concentrations

The effect of contact time and concentration on the ability of VCL and VCM to remove 4-aminoantipyrine from solution is presented in Fig. 7. The removal of 4-aminoantipyrine took place gradually as equilibrium was attained at 120 min for all concentrations above 1 mg/L. It was observed that as concentration reduced, the time taken to attain equilibrium also reduced. The relatively short time of attaining equilibrium (at 120 min) suggests that the sorption of 4-aminoantipyrine on both VCL and VCM may likely be via kinetics [36, 37]. For both VCL and VCM, the percentage removal of 4-aminoantipyrine increased as concentration reduced. However, the ability of VCL to remove 4-aminoantipyrine from solution was found lower than that of VCM. The capacity to remove 4-aminoantipyrine from solution improved significantly after the modification of VCL. This was observed in the increase in the percentage removal at concentrations higher than 1 mg/L in VCM. Vermiculites have high layer charge, which inhibits ion exchange. However, previous study has shown that acid treatment is capable of leaching the octahedral layers of vermiculites by way of protonating the oxygen atoms [38]. The treatment of VCL with NTA may have protonated the oxygen atoms in VCL, which may have resulted in increase in heterogeneity of OH species in the structure of VCM, and decreases layer charge, which may have promoted ion exchange. This change in structure must have enhanced the removal capacity of VCM toward 4-aminoantipyrine.

Fig. 7
figure 7

Percentage removal of 4-AAP by VCL and VCM at varying concentrations. Note: 4-AAP = 4-aminoantipyrine

Full size image

Adsorption kinetics

The adsorption kinetics of 4-aminoantipyrine on VCL and VCM was studied by fitting the experimental data with pseudo-first-order, pseudo-second-order, Elovich and intra-particle diffusion models. The linearized form of pseudo-first-order rate equation is [39]:

$$ {\text{In}}\left( {q_{\text{e}} - q_{\text{t}} } \right) = {\text{In}}q_{\text{e}} - k_{1} t $$
(3)

where qe and qt are concentrations of 4-aminoantipyrine at equilibrium and time t, respectively, while k1 is the pseudo-first-order rate constant (min−1). The equilibrium adsorption capacity (qe) and k1 were calculated from the intercept and slope of the plot of In (qe − qt) versus t.

The linearized form of the pseudo-second-order equation is [40]:

$$ \frac{t}{{q_{\text{t}} }} = \frac{1}{{k_{2} q_{\text{e}}^{2} }} + \frac{t}{{q_{\text{e}} }} $$
(4)

where k2 is the adsorption rate constant (g mg−1 min−1). The initial sorption rate was calculated as k2qe (mg g−1 min−1). Equilibrium adsorption capacity (qe) and the pseudo-second-order rate constant, k2, were obtained from the slope and intercept of the plot of t/qt versus t. The linearized form of Elovich equation was also used which is given as [41]:

$$ q_{\text{t}} = \frac{1}{\beta } {\text{In}}\left( { \propto \beta } \right) + \frac{1}{\beta }{\text{In}}t $$
(5)

There Elovich constants α (mg g−1 min−1) and β were calculated from the plot of qt versus In t. Data obtained were further fitted for the Intra-particle diffusion. This was determined as described by Weber and Morris [42]:

$$ q_{\text{t}} = k_{i} t^{0.5} $$
(6)

The rate constant, ki (mg g−1 min−0.5), was determined from the slop of the plot of qt versus t0.5.

The results for the different models described are shown in Table 1. The data fitted best for pseudo-second-order model, having better regression coefficients than other models considered. This suggests that the sorption of 4-aminoantipyrine on VCL and VCM follows pseudo-second-order kinetic model which indicates that the sorption process is chemically rate controlled suggesting chemisorption [43]. The adsorption capacity (qe) obtained for the pseudo-second-order kinetic model (2.7042 mg g−1 for VCL and 6.5317 mg g−1 for VCM) compared relatively well with values obtained from the experimental results for both VCL (2.9045 mg g−1) and VCM (6.5378 mg g−1). This adsorption capacity values are lower that values reported for modified organo-vermiculite for the removal of p-nitrophenol [44] and diatomite for the removal of quinmerac [45]. However, the percentage removal exhibited by VCL and VCM reveals them as substantial materials for the treatment of water system contaminated with low amount of 4-aminoantipyrine and possibly other organic micro-pollutant at low concentrations.

Table 1 Kinetic model parameters for the sorption of 4-aminoantipyrine on VCL and VCM
Full size table

The plot of qt versus ln t for the Elovich model gave a straight line from which β was calculated to be 2.0756 mg g−1 for VCL and 1.7879 mg g−1 for VCM. The r2 values were 0.9607 for VCL and 0.9195 for VCM which further supports the suggestion that the sorption of 4-aminoantipyrine on VCL and VCM may be via chemisorption [46].

Intra-particle diffusion further gave insight into the sorption of 4-aminoantipyrine on VCL and VCM. There is the possibility of the overall sorption process being controlled either by one or more steps. The possibility of intra-particle diffusion was examined using the intra-particle model. The plot of qt versus t0.5 gave a linear plot, which suggested the presence of intra-particle diffusion during the sorption process. However, the linear plot did not pass through the origin, which indicates that intra-particle diffusion was not the sole rate-controlling step [47].

Effect of pH

The pH plays a vital role in understanding the mechanism of sorption as this affects the charges at the surface of adsorbent as well as the nature of the adsorbate in solution. As shown in Fig. 8, the adsorption capacity of VCL and VCM increased with increase in pH. Similar observation was also recorded for the percentage adsorbed. This observed sorption pattern can be accounted for as being due to the surface charge of the adsorbents (VCL and VCM) and the adsorbate (4-aminoantipyrine) at the different pH range. The isoelectric point of vermiculite has been reported as 2.57 [23]. The surface of the adsorbents may be described as follows:

$$ - {\text{DOH}} + {\text{H}}^{ + } \rightleftarrows - {\text{DOH}}_{2}^{ + } \,\left[ {\text{Protonation}} \right] $$
(7)
$$ - {\text{DOH}} + {\text{OH}}^{ - } \rightleftarrows - {\text{DO}}^{ - } + {\text{H}}_{2} {\text{O}}\,\left[ {\text{Deprotonation}} \right] $$
(8)
$$ - {\text{DO}}^{ - } = - {\text{DOH}}_{2}^{ + } \,\left[ {\text{Isoelectric point}} \right] $$
(9)

where –D represents the surface of the adsorbents. In acidic medium (low pH), the surface can be described to be protonated as shown in Eq. 7. When in basic medium (high pH), the surface can be considered to be deprotonated as shown in Eq. 8 while the isoelectric point is represented in Eq. 9. The surfaces of the adsorbents were negatively charged at high pH which promoted the interaction with 4-aminoantipyrine which was positively charged at high pH. As seen in Fig. 8, the sorption of 4-aminoantipyrine by both adsorbents increased with increase in pH. Therefore, with increase in pH value, there was increase in the amount of 4-aminoantipyrine adsorbed by the adsorbents as a result of the increase in the electrostatic interaction between the surfaces of the adsorbents and 4-aminoantipyrine. However, the higher adsorption capacity exhibited by VCM over VCL may be due to the additional contribution from –COOH species on its surface resulting from the treatment with NTA. This additional contribution at higher pH may be described as:

Fig. 8
figure 8

Effect of pH on the removal of 4-aminoantipyrine by VCL and VCM

Full size image
$$ - {\text{DCOOH}} + {\text{OH}}^{ - } \rightleftarrows - {\text{DCOO}}^{ - } + {\text{H}}_{2} {\text{O}}\,\left[ {\text{Deprotonation}} \right] $$
(10)

This increases the negative charges at the surface of VCM, which further enhances its ability to remove 4-aminoantipyrine from solution. Moreover, treatment with NTA may have caused leaching of layer components and loss of CEC may be compensated by the observed increase in surface area and formation of adsorption centers or pores [27]; these adsorption centers may have also played a role in trapping some amount of 4-aminoantipyrine.

Adsorption isotherms

The experimental data were fitted with Langmuir, Freundlich and Temkin isotherm models. The Langmuir isotherm explains the existence of maximum limitation uptake, which reflects a saturated monolayer of 4-aminoantipyrine species at the surface of VCL and VCM with an assumption of uniform energies of adsorption. The linear representation of the Langmuir isotherm model is expressed as:

$$ \frac{{C_{\text{e}} }}{{q_{\text{e}} }} = \frac{1}{{Q_{\text{o}} }}C_{\text{e}} + \frac{1}{{K_{\text{L}} }} Q_{\text{o}} $$
(11)

Ce (mg/L) represents the equilibrium concentration of 4-aminoantipyrine, Qo (mg g−1) is the maximum monolayer coverage capacity, qe (mg g−1) is the amount of 4-aminoantipyrine adsorbed at equilibrium and KL (L mg−1) is the Langmuir isotherm constant. A linear plot was obtained when Ce/qe was plotted against Ce with a slope of 1/Qo and an intercept of 1/QoKL. The essential features of the Langmuir isotherm (RL) may be expressed as:

$$ R_{\text{L}} = \frac{1}{{1 + \left( {1 + K_{\text{L}} C_{\text{o}} } \right)}} $$
(12)

where Co is the initial concentration and KL is the constant related to the energy of adsorption (Langmuir constant). When RL > 1 adsorption nature is considered to be unfavorable, linear if RL = 1, favorable if 0 < RL < 1 and irreversible if RL = 0. RL was 0.0041 for VCL and 0.0446 for VCM indicating a favorable process, which was monolayer. As shown in Table 2, the higher value of KL in VCL than that of VCM shows a higher heat of sorption and stronger bond of sorption of 4-aminoantipyrine at the surface of VCL. This also means that VCL holds 4-aminoantipyrine stronger at its surface than VCM does.

Table 2 4-Aminoantipyrine sorption parameters for Langmuir, Freundlich and Temkin models
Full size table

Freundlich isotherm describes an empirical relationship between the concentrations of 4-aminoantipyrine at the surfaces of VCL and VCM. This can be expressed as:

$$ q_{\text{e}} = K_{\text{f}} C_{\text{e}}^{n} $$
(13)

The model assumes that adsorption occurs on heterogeneous surface via multilayer sorption mechanism and the amount of 4-aminoantipyrine adsorbed increase with concentration. Where qe (mg g−1) is the amount of 4-aminoantipyrine adsorbed at equilibrium, Kf (mg g−1) is the Freundlich isotherm constant, Ce (mg L−1) is the equilibrium concentration of 4-aminoantipyrine and n is the adsorption intensity. The values obtained for the Freundlich parameters are presented in Table 2. Considering the Freundlich model, adsorption capacity can be explained as a function of 1/n. Previous work has shown that when 1/n = < 1, the process can be said to be a normal adsorption, if 1/n = 1 then the partition between the liquid and solid phases are independent of concentration and if 1/n = > 1 the process can be considered to be a cooperative adsorption [48]. 1/n was found to be 1.2659 for VCL and 0.3026 for VCM, respectively; which suggests a cooperative and normal adsorption, respectively.

Temkin isotherm model suggests adsorbent–adsorbate interaction and a linear decrease in heat of sorption with surface coverage [49] which is given as:

$$ q_{\text{e}} = \left( {\frac{RT}{b}} \right){\text{In}}\left( {AC_{\text{e}} } \right) $$
(14)
$$ q_{\text{e}} = B{\text{In}}A + B{\text{In}}C_{\text{e}} $$
(15)

where b is the Temkin isotherm constant, A (L g−1) represents Temkin isotherm equilibrium binding constant and constant B (J mol−1) = RT/b is the constant related to heat of sorption. T is the absolute temperature (K) and R stands for gas constant (8.314 J mol−1 K−1). Plot of qe against ln Ce gave a straight line from which B and A were determined from slope and intercept. Values obtained are shown in Table 2. The isotherm equilibrium binding constant and constant related to heat of sorption were found higher in VCL than VCM which further corroborated the preference 4-aminoantipyrine had for VCL over VCM. The correlation coefficients (r2) obtained for the three tested models conformed to experimented data with Langmuir model fitting best with r2 value 1.0000. This suggests a monolayer adsorption process.

Thermodynamics of adsorption

Thermodynamic parameters for the sorption of 4-aminoantipyrine by VCL and VCM were examined and are presented in Table 3. Parameters determined included Gibb’s free energy change (∆Go), entropy change (∆So) and enthalpy change (∆Ho). The adsorption equilibrium constant (bo) was estimated from the expression:

$$ b_{\text{o}} = \frac{{q_{\text{e}} }}{{C_{\text{e}} }} $$
(16)
$$ \Delta G^{\text{o}} = - RT{\text{In}}b_{\text{o}} $$
(17)
$$ \Delta G^{\text{o}} = \Delta H^{\text{o}} - T\Delta S^{\text{o}} $$
(18)

R is the universal gas constant (8.314 J mol−1 K−1), qe is the adsorption capacity of VCL or VCM at equilibrium, Ce is the equilibrium concentration of 4-aminoantipyrine and T is the absolute temperature in Kelvin. ∆Ho and ∆So were obtained from the slope and intercept of the plot of ln bo against reciprocal of temperature (1/T). The percentage removal of 4-aminoantipyrine by VCL and VCM decreased as temperature increased. The values of qe obtained for VCL and VCM reduced as temperature increased while the values for the Gibb’s free energy increased with increase in temperature. The negative nature of ∆Go is an indication that the sorption of 4-aminoantipyrine by VCL and VCM is a spontaneous reaction. Results obtained for ∆Ho and ∆So are shown in Table 4. Both ∆Ho and ∆So are negative for VCL (∆Ho was − 57.177 kJ mol−1 and was ∆So − 0.179 kJ mol−1) and VCM (∆Ho was − 29.701 kJ mol−1 and was ∆So − 0.089 kJ mol−1). These negative values are an indication that the sorption process is spontaneous at low temperature but non-spontaneous at high temperature. The negative nature of ∆Ho further suggests the process to be exothermic while the negative values for ∆So is an indication of a stable configuration of 4-aminoantipyrine on the surfaces of VCL and VCM.

Table 3 ΔG, percentage removal and qe obtained at various temperatures for VCL and VCM
Full size table
Table 4 Thermodynamic parameters obtained from plot of ln bo versus 1/T for VCL and VCM
Full size table

Desorption

The desorption process evaluated the removal of 4-aminoantipyrine from the surface of VCL and VCM with the aim of reuse. This involves understanding the regeneration capacity of VCL and VCM. This was achieved using ethanol, distilled water, benzene, 0.5 M HCl, 0.5 M NaOH and acetone as solvents for removing 4-aminoantipyrine from the surface of VCL and VCM. The results obtained for desorption of 4-aminoantipyrine from VCL and VCM are presented in Fig. 9. 4-Aminoantipyrine was best desorbed from VCL (60%) and VCM (76%) using water. Regeneration capacity was found lower when benzene and acetone were used as solvent for VCL, whereas the least desorption capacity was obtained using acetone. Distilled water may be considered as a potential solvent for the solvent regeneration of VCL and VCM loaded with 4-aminoantipyrine.

Fig. 9
figure 9

Desorption of 4-aminoantipyrine from VCL and VCM

Full size image

Quantum chemical computations

Quantum chemical computational analysis helped in understanding and describing the adsorption of 4-aminoantipyrine on VCL and VCM in molecular terms. DFT may be applied in discussing the electronic structural changes that occurs during adsorption process. Presently, DFT was applied to describe a theoretical framework in identifying the contributions of 4-aminoantipyrine adsorption mechanisms and strength. The DFT was performed at B3LYP/6-31G level theory using Spartan 14.1 software. The geometry optimization was followed by modeling the molecular electronic structures and the distribution of frontier molecular orbitals in order to establish the reactivity of 4-aminoantipyrine as well as its active sites. Figure 10 presents the optimized geometry, HOMO density distribution, LUMO density distribution, Mulliken charge population analysis, total electron distribution, electrostatic potential map, local ionization potential map as well as the LUMO map of 4-aminoantipyrine.

Fig. 10
figure 10

Electronic properties of 4-aminoantipyrine. a Optimized structure of 4-aminoantipyrine, b HOMO density distribution of 4-aminoantipyrine, c LUMO density distribution of 4-aminoantipyrine, d Mulliken charge population analysis of 4-aminoantipyrine, e Total electron density of 4-aminoantipyrine, f Electrostatic charge of 4-aminoantipyrine, g Electrostatic potential map of 4-aminoantipyrine, h Local ionization potential map of 4-aminoantipyrine, i LUMO map of 4-aminoantipyrine. Note: C = carbon, H = hydrogen, O = oxygen and N = nitrogen

Full size image

The LUMO and HOMO are distributed over the molecule of 4-aminoantipyrine due to the presence of lone pairs of electron from nitrogen, and oxygen and the π-electrons in its molecule. Region of highest electron density (HOMO) is possible site at which electrophiles attack, such sites are the active or adsorption centers with the utmost energy to bond to the surface of VCL and VCM. However, the LUMO orbital of 4-aminoantipyrine accepts electrons from the orbital of the constituent transition elements (Ti and Fe) in VCL and VCM using antibonding orbitals to form feedback bonds [50]. The presence of these adsorption centers are capable of forming flat orientation of 4-aminoantipyrine molecules on the surfaces of VCL and VCM, which may have given rise to the surface coverage by the molecule of 4-aminoantipyrine. The nitrogen and oxygen atoms present in 4-aminoantipyrine are most probably the reactive site for its sorption on the surfaces of VCL and VCM. This further suggests the interaction between the surfaces of VCL/VCM and 4-aminoantipyrine to occur via donor–acceptor interactions among lone pairs of electron, π-electrons and the vacant d-orbital of the constituent transition metal in VCL/VCM. The higher adsorption capacity observed in VCM must have been due to addition interaction between the heteroatoms in 4-aminoantipyrine and the carboxyl functional groups on VCM. The computed Mulliken atomic charges are as presented in Table 5. As seen in Fig. 10, nitrogen, oxygen and a few carbon atoms are negatively charged. The more negatively charged the atoms of the adsorbed center, the more readily they are available to interact with surfaces of VCL/VCM and the easily such atoms donate their electrons to unoccupied orbitals in constituent transition metal in VCL/VCM [51]. The Mulliken atomic charges revealed that molecule of 4-aminoantipyrine possess atoms that are capable of donating electrons to the orbital of constituent transition metals of VCL/VCM for bonding. These electrons are also available for covalent bonding with functional groups at the surfaces of VCL and VCM.

Table 5 Molecular properties of 4-aminoantipyrine calculated using DFT at B3LYP/6-31G basis set level
Full size table

Reactivity of 4-aminoantipyrine may be related to frontier molecular orbitals. The higher HOMO energy of 4-aminoantipyrine may mean higher electron donating capacity to low energy empty molecular orbital of transition metals in VCL/VCM. On the other hand, LUMO energy is related to susceptibility of 4-aminoantipyrine molecule to nucleophilic attack and electron affinity. So electron-accepting capacity becomes stronger with lower LUMO energy. The energy gap (ΔE) was calculated as:

$$ \Delta E = E_{\text{LUMO}} - E_{\text{HOMO}} $$
(19)

The value of the energy gap is an indication of electronic interaction. The lower the value, the better the interaction and hence sorption of 4-aminoantipyrine from solution. The value obtained for 4-aminoantipyrine is 4.32 eV. This value is within the range previously reported for the sorption of plant extract on metal surface [52]; the value is higher than what was reported for methionine and phenylalanine [53] but lower than what was reported for Piper guineense extract [54]. Dipole moment also describes sorption process; when it is high, it is an indication that adsorption process is favored. The dipole moment for 4-aminoantipyrine was found to be 6.55 debye. Absolute hardness (η) is important in measuring the reactivity and stability of 4-aminoantipyrine. It was calculated as:

$$ \eta = \frac{{E_{\text{LUMO}} - E_{\text{HOMO}} }}{2} $$
(20)

Adsorption of 4-aminoantipyrine took place on surface of VCL/VCM at regions on 4-aminoantipyrine molecule with greatest softness and lowest hardness. Hard molecules have large energy gap while soft molecules does not. This makes soft molecules more reactive than hard molecules [55].

Conclusion

The wastewater purifying capacity of VCL was enhanced by way of chemical modification using CNCs and NTA via simple reaction to produce VCM. Both VCL and VCM were characterized and investigated as adsorbent for the removal of 4-aminoantipyrine from aqueous solution. After the modification, the BET surface area of VCL increased from 4 to 96 m2 g−1 in VCM. The adsorption capacity of VCM (6.5378 mg g−1) was found higher than that of VCL (2.9045 mg g−1). The sorption process of VCL and VCM was chemisorptive, monolayer and spontaneous. Quantum chemical computational analysis was also used to understand and describe the sorption of 4-aminoantipyrine on VCL and VCM in molecular terms. It was found that the LUMO and HOMO are distributed over the molecule of 4-aminoantipyrine due to the presence of lone pairs of electron from nitrogen, and oxygen and the π-electrons in its molecule which aid its sorption on VCL and VCM. This interaction between the surfaces of VCL/VCM and 4-aminoantipyrine occurred via donor–acceptor interactions among lone pairs of electron, π-electrons and the vacant d-orbital of the constituent transition metal in VCL/VCM. The results suggest VCM as a potential resource for the removal of pharmaceutical pollutants from water system.