Functionalized MWCNTs-quartzite nanocomposite coated with Dacryodes edulis stem bark extract for the attenuation of hexavalent chromium

Multiwalled carbon nanotubes/quartzite nanocomposite modified with the extract of Dacryodes edulis leaves was synthesized and designated as Q, which was applied for the removal of Cr(VI) from water. The adsorbents (PQ and Q) were characterized using the SEM, EDX, FTIR, TGA, XRD, and BET analyses. The XRD revealed the crystalline composition of the nanocomposite while the TGA indicated the incorporated extract as the primary component that degraded with an increase in temperature. The implication of the modifier was noticed to enhance the adsorption capacity of Q for Cr(VI) by the introduction of chemical functional groups. Optimum Cr(VI) removal was noticed at a pH of 2.0, adsorbent dose (50 mg), initial concentration (100 mg dm−3), and contact time (180 min). The kinetic adsorption data for both adsorbents was noticed to fit well to the pseudo-second-order model. The adsorption equilibrium data were best described by the Langmuir model. The uptake of Cr(VI) onto PQ and Q was feasible, endothermic (ΔH: PQ = 1.194 kJ mol−1 and Q = 34.64 kJ mol−1) and entropy-driven (ΔS : PQ = 64.89 J K−1 mol−1 and q = 189.7 J K−1 mol−1). Hence, the nanocomposite demonstrated potential for robust capacity to trap Cr(VI) from aqueous solution.


Determination of pH point of zero charge (pH PZC ).
To determine the pH at the point of zero charges of PQ and Q, about 0.1 g of the materials was transferred into eleven 250 cm 3 glass Erlenmeyer flasks with each containing 50 cm 3 of 0.1 mol dm −3 NaCl solution at pH values ranging from 2 to 12. The flasks were stoppered and agitated for 48 h in a preset thermo-regulated water bath at 25 °C. The final pH of the mixture was obtained after 48 h and a plot of the final pH versus the initial pH was made from which the pH PZC of PQ and Q was extrapolated from the line intercept 55 .
Batch adsorption experiments. Batch  To assess the thermodynamics of the adsorptive process, the initial concentration experiment was repeated for 303 K, 308 K and 313 K. The residual concentration of Cr(VI) was estimated by the colorimetric method using 1,5-diphenyl-carbazide as a complexing agent, analyzed using the UV-visible spectrophotometer at 540 nm 56 .
The adsorption capacities and the uptake efficiency of PQ and Q were calculated as shown in the supplementary information.
Kinetics and isotherm models. The kinetic modeling of adsorption was performed using four kinetic models while the isotherm analysis was carried out by applying eight isotherm models as described in the supplementary information.

Reusability experiments.
To examine the reusability of PQ and Q, the adsorbents were used to remove the adsorbate (Cr(VI)) from the aqueous phase using the same adsorption procedure stated in the previous section. The regeneration of PQ-Cr and Q-Cr was performed by making use of NaOH. About 0.5 g of PQ-Cr or Q-Cr was in contact with 25 cm 3 of 0.5 mol dm −3 NaOH for 3 h at 25 °C. The regenerated adsorbents were washed twice with ultrapure water and dried for the next cycle. The removal efficiency of PQ and Q for the next cycle was estimated using Eq. (2).
Compliance with ethical standard. The collection of plant materials complied with relevant institutional, national, international guidelines and legislations.

Results and discussion
Characterizations of adsorbents. The micrographs of the surface morphology of PQ and Q are displayed in Fig. 1. The SEM micrograph of the pristine quartzite (PQ) revealed an aggregate of smooth particles with varied shapes and sizes (Fig. 1a). The structure of PQ reflected the metamorphic nature of quartzite rock. The micrograph (Fig. 1b) of the nanocomposite(Q) showed that the quartzite particles were rapped with an intertwined network of cylindrical tubes resulting in the formation of microscopic channels on the surface with an increase in surface roughness probably due to coating by the incorporated extract. The micrograph (Fig. 1c) of the nanocomposite adsorbed with Cr(VI) (Q-Cr) was acquired at a higher magnification which further revealed an increase in surface roughness of the grains without the segregation of the network of carbon nanotubes or comparable sign of degradation concerning the surface integrity of the nanocomposite. This indicated the stability of the material in the aqueous environment. To further confirm the adsorption of chromium onto the surface of Q, the acquired EDX of Q-Cr (Fig. 1d) revealed the surface adsorbed chromium. It should be noted that the percentage (0.47%) indicated by EDX represented Cr on the spot being scanned and not the total adsorbed Cr on the nanocomposite. X-ray diffraction analysis was used to assess the mineral content of quartzite. As shown in Fig (Fig. 1d) showed consistency with the identified minerals in the diffractogram. The diffractogram of the nanocomposite (Q) identified the same mineral with a reduction in the intensities of the peaks which could be associated with the surface coating and the presence of the carbon nanotubes. However, the diffraction patterns associated with the carbon nanotubes were not observed, which indicated the dominance of the quartzite mineral intensities due to higher crystallinity. www.nature.com/scientificreports/ The FTIR spectra of PQ and Q were recorded in the wavenumber range of 4000-400 cm −1 . The prominent peaks in the spectrum of PQ (Fig. 3) appeared in the range 1000-400 cm −1 . These peaks were associated with quartz, which indicated Si-O-Si 60 . However, additional bands were observed in the spectrum of the nanocomposite at (ν/cm −1 ): 1380, 1610, 3500 assigned to asymmetric vibrational stretching modes of carboxylate group (-COO-), the symmetrical C=O stretching of ionic carboxylate groups and the -OH stretching [61][62] . These peaks indicated the incorporation of the Dacryodes edulis leaves extract as well as the functionalized carbon nanotubes on the surface of the adsorbent. In a bid to assess the interaction of the adsorbate on the surface of the adsorbents, the FTIR spectra of the loaded adsorbents (PQ-Cr and Q-Cr) were acquired. As shown in Fig. 3, the intensity associated with the band of the hydroxyl functional group on the surface of the nanocomposite was reduced after the adsorption step. This suggested that the -OH functional groups were actively involved in the adsorption of Cr(VI) onto the nanocomposite.
The surface area and pore volume of PQ and Q were assessed using the BET nitrogen adsorption-desorption technique. Meanwhile, the pore diameter of PQ and Q were estimated using the Barrett-Joyner-Halenda (BJH) approach. The physisorption isotherm of PQ indicates poor adsorbate-adsorbent interaction. However, the incorporation of carbon nanotubes and plant extract from Dacryodes edulis showed a distinct improvement in the specific surface area of the nanocomposite over PQ (Fig. 4). This indicated an increase in the rate of interaction due to the availability of more surface area. Similarly, the increase in pore diameter and pore volume (Table 1) of the nanocomposite (Q) compared to PQ indicated the introduction of macropores along with the existing mesoporous structure of PQ. The form of the graph of amount gas adsorbed versus partial pressure is known as an adsorption isotherm, which are classified as type I-V according to their shape 63 . A type III physisorption isotherm was observed for PQ while a type IV isotherm was observed for the nanocomposite, which indicated the enhanced porosity of the nanocomposite.
The thermal stability of Q was investigated using the thermogravimetric analysis technique. The thermal behavior of pristine quartzite as reported by Xing et al.revealed the loss of physisorbed water only 64 . However, the presence of plant extract and carbon nanotubes strongly influence the thermal stability of the nanocomposite. Figure 5, displayed the thermogram of the nanocomposite which revealed a unique decomposition pattern. www.nature.com/scientificreports/ About 1.90% mass loss was noticed with the rise of temperature to 110 °C. This could be attributed to the loss of physisorbed water. A further reduction of 2.06% observed as the temperature increased to 300 °C could be associated with the loss of decomposed plant extract on the surface of the nanocomposite and the loss of internally bonded water. However, a progressive decline in mass was observed in the temperature range of 300-800 °C, which reflects the elimination of volatile inorganic or organic materials that were internally bonded. Over the investigated temperature range, about 17.42% mass loss was noticed for the nanocomposite. This accounted for the incorporated plant extract and it showed the stability of quartzite framework, which provided support for the carbon nanotubes used in the fabrication of the nanocomposite. Hence, the fabricated adsorbent can be used to treat wastewater even at high temperatures (< 300 °C) without loss of morphological integrity.
Effect of pH on Cr(VI) uptake. Figure 6 revealed the implication of varying sorbate pH on the adsorptive removal of Cr(VI) by PQ and Q. The optimum removal capacities of PQ (27.50 mg g −1 ) and Q (45.87 mg g −1 ) were observed at pH 2. It is noted that at all pH values, the surface modification of PQ to give Q, led to about twice as much removal capacity of Q than for PQ. Meanwhile, the adsorptive capacity of PQ and Q was noticed to decrease with decreased acidity of Cr(VI) solution. Hence, the nanocomposite (Q) demonstrated an adequate potential to eliminate hazardous hexavalent chromium from the aquatic ecosystem. Speciation of hexavalent chromium and the pH PZC of the adsorbents offers better insight on the high Cr(VI) uptake capacity of the sorbent at low solution pH. Note that a reducing argent is necessary to reduce Cr(VI) to Cr(III) at a low pH. As shown in Fig. 7, the pH PZC of PQ and Q were 4.08 and 6.18, respectively. This indicates that at solution pH below and above these values, the surface of the adsorbents (PQ and Q) will be positively and negatively charged respectively. Hence, at solution pH 2, the adsorbents will be positively charged. The hexavalent chromium exists in different oxyanions form with variation in solution pH. In the pH range of 2 to 4, Cr(VI) exist mainly as HCrO 4 in an aqueous solution. Meanwhile, as the pH increase from 4 to 6, HCrO 4 − is converted to and in equilibrium with Cr 2 O 7 2- [65][66] . The enhanced uptake capacity of PQ and Q at pH 2 could be due to electrostatic interaction between the positively charged surface of the adsorbents and negatively charged chromium species. It is also possible that Q acts as a reducing agent during the Cr(VI) uptake process, due to the reducing potential of Dacryodes edulis 54 . However, finding from this study is in good agreement with the reports from other authors that got optimum pH of 2 22,67 .   www.nature.com/scientificreports/     Fig. 8, the time-dependent adsorptive removal of Cr(VI) by PQ and Q was in three stages. The first stage was a fast phase (occurred before 20 min), followed by a gradual increase until 180 min and the last stage that involved no significant increase in the uptake capacity of PQ and Q. Hence, 180 min was selected as the optimum contact time for the elimination Cr(VI) from aqueous solution. However, to ensure complete removal of the adsorbate, 1440 min was employed for further experiment. PQ adsorbed more than double the amount that Q adsorbed at all times > 20 min.

Kinetics study.
To shed light on the Cr(VI) removal rate by PQ and Q, kinetics model namely pseudo-firstorder, pseudo-second-order, Elovich, and Moris-Weber intraparticle diffusion were employed. The nonlinear equations of these models were used to analyze the kinetic data (see Table 2) and the acquired plots are shown in Fig. 9. Meanwhile, the least sum of square residuals (SSR) was used as the goodness-of-fit measure in selecting the model that best describes the data ( Table 2). The model with the smallest SSR value gave the best fit to  www.nature.com/scientificreports/ the experimental. The pseudo-first-order kinetic model is best used to describe the adsorption process that is dominantly controlled by diffusion. This model is also based on the fact that an adsorbate binds to a single adsorption site. On the other hand, an adsorption process that is driven by chemisorption is best described by pseudo-second-order kinetic models. In principle, the pseudo-second-order kinetic model assumes binary adsorption of a sorbate. An adsorptive model that demonstrates chemisorption as its rate-determining step is described by the Elovich kinetic model. However, Morris-Weber intraparticle diffusion is used to describe the basic stages involved in the adsorption process. These stages include (i) conveyance of sorbate from the solution bulk to a thin film layer, (ii) transport of sorbate from film layer to the surface of the adsorbent, (iii) migration of sorbate from the adsorbent surface to the interior of the porous structure and (iv) adsorption of sorbate to the adsorption sites. As shown in Table 2, the adsorptive removal of Cr(VI) by PQ and Q was best described by pseudo-second-order (SSR = 10.87) and Elovich kinetic models (SSR = 32.32) respectively. This suggested that chemisorption was the rate-limiting step in the adsorptive removal of hexavalent chromium from the aquatic solution.
Effect of adsorbent dose. As shown in Fig. 10, an increase in removal efficiencies of PQ (15.4 to 67.7%) and Q (73.1 to 98.7%) was noticed as the adsorbent dose was increased from 0.01 to 0.4 g (see Fig. 10). This could be attributed to the increasing amount of adsorption sites at fixed initial Cr(VI) concentration. On the contrary, the uptake capacity of PQ (38.5 to 4.2 mg g −1 ) and Q (182.7 to 6.2 mg g −1 ) were observed to decrease with increased dosage. The decrease in the uptake capacity could be associated with the agglomeration of the sorbent at a higher dosage. Irrespective the adsorbent dose, Q outperformed PQ by having about double the removal capacity of PQ.

Effect of initial concentration and solution temperature. The implication of initial concentration
on the adsorptive removal of Cr(VI) by PQ and Q was investigated and the results are displayed in Fig. 11. This  www.nature.com/scientificreports/ study was examined over a concentration range of 10-100 mg g −1 with a fixed adsorbent dose of 0.25 g. As shown in Fig. 11, the result revealed that the uptake capacities of PQ and Q increased from 2.5 to 24.3 mg g −1 and 5.1 to 44.0 mg g −1 respectively as the initial Cr(VI) concentration increased from 10 to 100 mg dm −3 at 298 K. It indicated that the adsorptive removal of Cr(VI) onto PQ and Q was strappingly dependent on the initial hexavalent chromium concentration 68 . The phenomenon could be associated with enhanced collisions frequency between the Cr(VI) ions and the adsorption sites on the surface PQ and Q, resulting in higher surface coverage at high initial Cr(VI) concentrations, and thus high uptake capacities. A similar trend was observed at all temperatures investigated. However, an increase in solution temperature was noticed to enhance the uptake capacity of Q, although this effect was more significant at higher solution temperature and thus, demonstration the endothermic adsorptive process. In contrast to this, the influence of solution temperature on the adsorption of hexavalent chromium onto the surface of PQ was trivial. The quantity of Cr(VI) adsorbed at equilibrium, q e , is ca double for PQ compared to the Q at the same initial concentration of Cr(VI).
Adsorption isotherm. The distribution of sorbate in the solid-liquid interphase and the estimation of sorbent potential to eliminate specific sorbate can be determined by making use of mathematical models termed adsorption isotherm. To understand the mechanism of Cr(VI) adsorption onto the surface of PQ and Q, various two-and three-parameter isotherms were used to analyze the equilibrium adsorption data. The least sum of squared residuals (SSR) and the residual squared errors (RSE) were used to select the model that best describes the experimental data. A smaller sum of squared residuals value indicate a better fit of the model to the data; a value of zero means a perfect fit of the model. Among the two-parameter models (Dubinin-Radushkevick, Temkin, Freundlich, Langmuir), the Langmuir isotherm was adequate to describe the adsorption of Cr(VI) onto PQ and Q within the studied temperature range observed (see Table 3). This suggested that the adsorption process of Cr(VI) by PQ and Q was mainly monolayer adsorption. However, those of the three-parameter models (Toth, Redlich-Peterson, Khan, Sips), Sips and R-P were noticed to best fit PQ and Q respectively.

Comparison of adsorbents for Cr(VI).
A comparison of the maximum monolayer capacity (q max ) of PQ and Q, with previously reported adsorbents, showed that the nanocomposite reported in this study possess a better absorbance capacity for application in environmental remediation practice (see Table 4). Especially is noted that the surface modification of PQ led to about five times the maximum removal capacity for Q, compared to PQ.
Adsorption thermodynamics. Evaluation of the thermodynamics of adsorption was performed as decribed in the supplementary information. Table 5 revealed that at all temperatures investigated, the ∆G° values of the adsorption of Cr(VI) onto PQ and Q were negative. It showed that the uptake of Cr(VI) by PQ and Q were feasible and spontaneous. Meanwhile, a slight increase in the ∆G° values was noticed with increased solution temperature. This indicated that the removal efficiency of the adsorbents was favored at higher solution temperatures. Furthermore, positive values of ΔH° and ΔS° were estimated from the thermodynamic analysis. This indicated an endothermic adsorption process and an increased haphazardness at the sorbate-sorbent interface respectively. It is worth mentioning that the thermodynamic of an adsorption process creates a path to unveiling the adsorptive mechanism. However, literature revealed that the adsorption process with ΔH° values between 2.1 and 20.9 kJ mol −1 is physisorption [77][78] , while ΔH° values are between 80 to 200 kJ mol −1 is chemisorption driven 77 . Hence, regarding the ΔH° values of PQ and Q as display in Table 4, the adsorptive removal of Cr(VI) by PQ was a physisorption process while the uptake of Cr(VI) onto Q was a physicochemical process 1 .

Reusability of PQ and Q.
The tendency of PQ and Q to retain their adsorption efficiency after multiple usages were examined and displayed in Fig. 12. To achieve this, the adsorption-desorption cycle was repeated five times. After the fifth cycle, PQ and Q were noticed to have an efficiency of 29% and 78%. The high reusability of Q after 5 cycles is an indicator of the excellent design of Q for Cr(VI) absorbance from waste water. This Figure 11. The effect of initial concentration on uptake of Cr(VI) by PQ and Q.    Adsorption mechanism. The mechanism responsible for the adsorption of Cr(VI) uptake onto Q may be due to the surface modification of MWCNTs-quartzite using Dacryodes edulis leaves extract. The optimum uptake of Cr(VI) onto PQ and Q was established at pH2 (see Fig. 6). Meanwhile, the pH PZC of PQ and Q were 4.08 and 6.18 respectively (see Fig. 7). Hence, at pH2 the positively charged surface of PQ and Q would interact with oxyanions species of Cr(VI) electrostatically. The functional groups of the phytochemical constituent of most plant extract, confer reductive characteristics on extract obtained from different part of plant. [REF].
Hence, the reduction of Cr(VI) to Cr(III) followed by the adsorption of Cr(III) onto the surface of Q via electrostatic interaction may be a possible route for the uptake of Cr(VI). The uptake of Cr(III) could also be via pore entrapment (the enhanced pore volume, pore dimeter and surface area of the nanocomposite (Q) could aid pore entrapment, see Table 1). As shown in Fig. 3, the FTIR spectra of the spent adsorbent (Q-Cr) revealed the disappearance of -OH bands, demonstrating that Cr(VI) chemically interacted with the surface of Q via chemisorption. The pseudo second order model was observed to best describe the uptake of Cr(VI) on to Q (see Table 2), This further justified the inclusion of chemisorption in the uptake of Cr(VI) onto Q.

Conclusion
A newly fabricated nanocomposite adsorbent (Q) prepared from multiwalled carbon nanotubes (MWCNT) and quartzite (PQ) coated with plant extract was assessed for its capacity to trap/reduce Cr(VI) to its nontoxic species from an aqueous solution. The nanocomposite (Q) showed good surface morphology, undisrupted crystalline phases, enhanced thermal stability, improved pore dimeter and enhanced surface area. Meanwhile, the optimum adsorptive conditions for removal of chromium(VI) by Q was established to be pH 2, adsorbent dose of 50 mg, initial chromium(VI) concentration of 100 mg dm −3 , the temperature of 318 K, and a contact time of 180 min. Under this conditions Q gave a maximum monolayer capacity (q max ) of 192.5 mg g −1 . Furthermore, the experimental isotherms data obtained for the uptake of chromium(VI) onto PQ and Q were best described by the Langmuir model. The time-dependent adsorption data were best described by pseudo-second-order kinetic. The estimated thermodynamic parameters suggested that the removal of chromium(VI) by PQ and Q was spontaneous, endothermic, and entropy-driven. The adsorbent (Q) has demonstrated robust efficiency and large absorbance capacity for the removal of chromium(VI) from aqueous solutions. Owing to the excellent reusability of the adsorbent, Q, giving 78% Cr(VI) absorbance after 5 cycles, Q are recommended to be tested scaled-up conditions for industrial applications.