Polycyclic Aromatic Hydrocarbons Removal from Aqueous Solution with PABA-MCM-41/Polyethersulfone Mixed Matrix Membranes

Polycyclic aromatic hydrocarbons (PAHs) are one of the most recalcitrant pollutant originated from the burning of coal, petroleum, and other fossil fuels. The human exposure to PAHs may contribute to develop several carcinogenesis mechanisms. The aim of the present study was to develop a mixed matrix membrane (MMM) based on polyethersulfone (PES) and functionalized mesoporous material for the remediation of PAHs mixture by adsorption processes. MCM-41-based mesoparticles were obtained from biomass reuse of rice husk ash (RHA) and functionalized with p-aminobenzoic acid (PABA). The hydrothermal and casting methods were effective and sustainable in the preparation of PABA-MCM-41 and PES-based MMMs, respectively. PES-based MMMs presented an excellent distribution of the arrays incorporated and small-angle ordering. The absorption of PAHs was influenced by the incorporation of PABA-MCM-41 within the PES matrix. Preparation of mixed matrix membrane of PES/PABA−MCM − 41. Preparation of mixed matrix membrane of PES/PABA−MCM − 41.

Recently, alternatives have been sought to improve the performance of conventional polymer membranes, this from the incorporation of nanoparticles into the polymer matrix. Thus, this new trend is called mixed matrix membranes (MMMs), which combine the properties of polymer matrix with the filler material. [6,30]. Thus, the performance of MMMs can be influenced by different factors, mainly: (i) characteristics of polymer and architecture of the filler material, (ii) compatibility, and (iii) preparation technique of MMMs [31].
In this sense, membranes can play a crucial role in the separation process of emerging contaminants. Among the main organic pollutants that are released in the environment are polycyclic aromatic hydrocarbons (PAHs) [35,39,41]. PAHs are mainly from the exploration of petroleum, combustion of fossil fuels, and automobile exhaust emissions, which are classified as non-biodegradable, mutagenic, allergenic, teratogenic, toxic, and/or carcinogenic compounds by the International Agencies [44,45].
In this investigation, modified mesoporous structure prepared from RHA was applied as filler array in polyethersulfone (PES) MMMs to be tested in the PAHs remediation from aqueous media. The MCM-41-based mesoparticles obtained from biomass were modified with PABA group. Our research group has demonstrated that PABA-MCM-41 is a potential good absorbent for PAHs [35,46]. From our knowledge, there are few investigations on the use of functionalized mesoporous materials in MMMs for PAHs remediation from aqueous media. Consequently, the present work can contribute significantly to the evaluation of a low-cost and eco-friendly biomass-based membrane for field of wastewater treatment.

Synthesis of the PABA-MCM-41 (RHA) Functionalized Mesoporous Material
The synthesis of functionalized mesoporous material was performed previously by our group from the hydrothermal/cocondensation method [46].

Preparation of MMMs
The membranes were made according to the methodology reported by Khan et al. [48], with modifications: PES and PABA-MCM-41 mesoporous material were dried at 85 and 150°C for 2 h, respectively. A solution of 20 wt.% PES was prepared in chloroform under constant stirring for 24 h. The resulting solution was casts on a petri dish and covered by an inverted glass funnel, this procedure was performed in more than one petri dish until the total use of the PES solution. At the end of 24 h, the funnel was removed to evaporate the residual solvent at room temperature for another 24 h. The pure PES membrane was then dried for 12 h (85°C) in a vacuum oven to further remove the residual solvent.
The preparation of the PES/PABA-MCM-41 MMMs adopted the following procedure: amounts of the PABA-MCM-41 in range 1-10.0 wt.% were dispersed under magnetic stirring in the PES solution for 24 h at room temperature. Finally, the same procedure to obtaining the pristine PES membrane was performed.

Transport Properties of MMMs
In this section, all tests were performed in duplicate.

Water Static Sorption
The assays were realized in ultrapure water, in which the PESbased MMMs were weighed and submerged in water until to 15 days, after which they were weighed again, and the swelling percentage was assessed from the expression: where m i and m f are the initial and final masses (g), respectively.

Permeation/Absorption of the PAHs
The permeation assessment was performed at room temperature using a side-by-side glass diffusion cell with two independent compartments separated by the PES-based MMMs. The PES-based MMMs were introduced between the two compartments of the cell and the system was sealed (constant pressure of 1.1 MPa, and membrane contact area between the glass cells of~2.31 cm 2 ). All permeation assays were per- The results found were shown in terms of the permeation rate (PR), calculated using the expression: where PR is the permeation rate (%), C F,0h is the initial concentration of PAHs of feed phase (μg L −1 ) and C P,48h is the PAHs concentration of permeate phase after 48 h (μg L −1 ).

Determination of B[b]F, B[K]F, and B[a]P PAHs
Polycyclic aromatic hydrocarbons were quantified on an HPLC using a Shimadzu LC-20A Prominence equipment according to the methodology developed in our research group [41], and the r 2 values obtained were 0.

Characterizations
The ATR-FTIR spectra of PES-based MMMs were measurement in a Varian 3100 spectrophotometer (at room temperature, 4000-600 cm −1 region, resolution of 2 cm −1 , and 20 scans). The materials were analyzed by XRD in a Shimadzu LabX XRD-6000 diffractometer operating with Cu Kα (λ = 1.5406 Ǻ, room temperature, 2θ range of 5-50°, width of 0.02°, step time of 0.6 s, and a scanning rate of 2°min −1 . SAXS assays were done according to the previously reported methodology [35]. SEM was required in a Phillips FEGXL 30/EDS equipment (3 kV). DSC analyzes of the PES-based MMMs were realized in a NETZSCH DSC 200 F3 apparatus (50 mL min −1 flow of nitrogen). Heating and cooling cycles between 20 and 240°C (20°C min −1 ) and the glass transition temperature (Tg) of PES-based MMM was determined in the second heating by the inflection point method.

Characterizations
The characterizations of the RH and RHA materials are reported previously [47,50]. The PABA-MCM-41 (RHA) modified mesoporous array exhibited a hexagonal arrangement with surface area value of the magnitude of 438 m 2 g −1 , the N 2 adsorption/desorption analysis also revealed a type IV isotherm with H1 hysteresis. Finally, the mesoporous structure showed an excellent thermal stability at a temperature below 177°C with total pore volume and pore diameter values of 0.41 cm 3 g −1 and 3.59 nm, respectively [46]. In addition, the PABA-MCM-41 (RHA) structure presented the typical bands of the modifying groups (PABA-Si) within the mesoporous matrix, as demonstrated previously [46]. Fig. 1(i) shows the ATR-FTIR spectra obtained for the pristine PES membrane and the MMMs prepared with the PABA-MCM-41 (RHA) in percentages between 1.0-10.0 wt.%. It is possible to observe the main characteristic bands of the PES, which are centered in the range between 2962 and 2839 cm −1 and those at around 831, 799, and 689 cm −1 are assigned to the stretching and vibration of C-H bond of the backbone of the aromatic hydrocarbon of PES, respectively [51,52]. The band centered at 1408 cm −1 is assigned to the −CH 3 bond of the CH 3 − C − CH 3 group [52,53]. The band at 1231 cm −1 and those between 872 and 856 cm −1 are attributed to the asymmetric vibration of the ether group [51,53].
The bands around 1321, 1296 and 1009 cm −1 can be assigned to the symmetrical and asymmetric O=S=O vibrations of the sulfone group, as well as the bands near 1145 and 1100 cm −1 are attributed to the S=O elongation. The bands around 1577 and 1483 cm −1 belonging to the C=C bond of the aromatic groups of PES [51,54,55].
The spectra of PES-based MMMs show an increase of the band intensity between 2962 and 2839 cm −1 . These bands are attributed to the vibrations of the methylene groups (−(−CH 2 ) 3 −) of the PABA-Si present into the PABA-MCM-41 functionalized mesoporous material [39,46], as well as the appearance of a band at around 1725 cm −1 on the MMMs, which can be assigned to the C=O stretching of carboxylic acid present on the PABA-Si functional group from PABA-MCM-41 (RHA). Likewise, the new band close to 1535 cm −1 is assigned to the amide −CONH− vibrations of PABA-Si groups [39], confirming that there was a good dispersion of mesoparticles on the PES MMMs.
The presence of the main bands of PES on the MMMs confirms the compatibility between the mesoparticles and the polymer chains. In this way, a decrease in the intensity of the bands close to 1321, 1296, and 1009 cm −1 (O=S=O symmetric and asymmetric vibrations), 1145 and 1100 cm −1 (elongation of S=O bond), and the bands centered at around 1577 and 1483 cm −1 (C=C bond of aromatic groups of PES) with the increase of the PABA-MCM-41 amount. Therefore, this decrease is attributable to the interaction between these PES groups with the carboxylic acid groups (−CO 2 H) present within the PABAMCM-41 arrays. However, the characteristic bands of amorphous silica of the PABA-MCM-41 were not evidenced in the spectra of the MMMs because of the overlapping of these bands in the presence of PES. Fig. 1(ii) shows the X-ray diffractograms obtained of PABA-MCM-41 (RHA), pristine PES membrane, and MMMs. The PABA-MCM-41 and the PES membrane exhibited a single diffraction peak, which is attributed to the amorphous halo of the silica source used on the PABA-MCM-41 and the amorphous structure of PES, respectively. It is also possible to observe that the MMMs presented the same wide band observed for the pristine membrane. This behavior is due to the compatibility between PABA-MCM-41 mesoparticles and the PES matrix. Fig. 1(iii) shows the SAXS profiles obtained of pristine membrane and MMMs. The PABA-MCM-41 array exhibited a ratio of relative distances of 1:√3:2, derived from the peaks ratio (q 100 /q 100 , q 110 /q 100 , and q 200 /q 100 ), respectively, confirming the mesoporous pattern of PABA-MCM-41 [46]. However, the pristine membrane did not present SAXS profile. Conversely, the PES-based MMMs presented three peaks at around the scattering vector q 100 , q 110 , and q 200 related to the (100), (110), and (200) planes from the PABA-MCM-41 within the MMMs. Furthermore, there was an increase in peak intensity q 100 with increase in the proportion of mesoparticles within the MMMs. Fig. 1(iv) shows the DSC curves of the second heating obtained of the pristine membrane and MMMs, while Table 1 shows the glass transition temperature (Tg) values found for same samples. There was only one thermal event for all samples, which is associated with the glass transition, where it is possible to observe a small decrease in Tg values of the PES-based MMMs compared to pristine membrane. The presence of the mesoparticles within the PES matrix seems to   Figure 2 shows the SEM images obtained of the pristine membrane and PES-based MMMs. The pristine membrane (Fig. 2a) exhibits an asymmetrical structure with a dense layer at top and a sublayer in the form of elongated channels. In contrast, the MMMs (Figs. 2b-f) show the appearance of interfacial voids surrounded by cavities at around PABA-MCM-41 (RHA) mesoparticles incorporated within the PES matrix, which increased with increasing of the mesoparticle content.
The findings of the chemical compositions of the pristine membrane and PES-based MMMs obtained from the SEM-EDS analyzes are shown in Table 2. The C contents did not have significant changes for the MMMs compared to pristine membrane, the same behavior was seen for the impurity contents of Cr, Al, Fe, and Ca. On the other side, the Si and O contents increased with the increase in the percentage of PABA-MCM-41, but the same was not noticed for S and Cl contents.

Absorption of MMMs
The WVT and P values for the pristine PES and MMMs are shown in Table 3. The WVT values for the MMMs were higher than the value obtained for the pristine membrane. The same behavior was also observed for the P values, except in MMM with 1.0 wt.% of mesoparticles. This phenomenon occurs due to the increase in the number of preferred paths for the water vapor transport with the incorporation of the mesoporous material within the MMMs, as seen in the SEM results. Figure 3(a) shows the correlation between the P and Tg values of the MMMs. It is possible to observe an inverse relationship between P and Tg values, except for MMM with 1.0 wt.%. This behavior occurs because at low Tg values, there is a greater mobility of the polymer chains, thus facilitating the permeation process through these MMMs. Table 4 presents the water static sorption values for the pure PES membrane and MMMs. There was a decrease in the swelling degree after incorporation of 1.0 wt.% of PABA-MCM-41 in the PES matrix. This behavior can be related to the occupation of the PES free volume by the well dispersed mesoparticles. However, for the 2.5, 7.5, and 10.0 wt.% contents, the swelling degree values were higher than those obtained for the other MMMs, and this can also be related to the occupation of the water molecules in the pores of the mesoparticles within the MMMs. Figure 3(b) shows the correlation of the swelling degree with the Tg values for the MMMs. There is a direct correlation between the mentioned values, so an increase in Tg causes an increase in the amount of water imprisoned in the free volume of the polymer matrix, as well as in the pores of PABA-MCM-41. In contrast, this behavior reflected in lower P values (Fig.  3a), because of the strong imprisonment of the water molecules by the silanol and PABA-Si groups of the modified mesomaterial structure within the PES-based MMMs. Figure 4(a-c) shows the results found from the PAHs concentration in feed solution against time for PES-based MMMs. The increase in the operation time caused a decrease in the PAHs concentrations since the equilibrium time found was at around 24 h for all MMMs. As with the final PAHs concentrations of the pristine membrane, whose values were lower than those obtained for the PES-based MMMs.
In contrast, Fig. 4(d-f) shows the PAHs concentration values of permeated solution through the MMMs. An increase in the permeate concentration was observed as the operating time was longer, given that the equilibrium time found was in the order of 24 h. The values of the final concentrations permeated through the pristine membrane, for all PAHs, were lower than those obtained for the PES-based MMMs. However, these values are close to those found for the MMM with a content of 7.5 wt.% mesoparticles. In view of these results, we can suggest that the occupation of the free volume of PES by the mesoparticles facilitated the permeation of PAHs through these MMMs, due to the appearance of interfacial voids within the MMMs, as observed in the SEM analyzes. Figure 4(g-i) shows the results of adsorbed concentrations of PAHs mixture by the pure PES membrane and MMMs. It is The results of permeation (PR) and retention (RR) rates and the removal percentage of PAHs mixture by PES-based MMMs were compiled in Table 5. The PR values observed for the PAHs mixture followed the following increasing order: [a]P, due to the hydrophobic effect of PAHs mixture in solution. Therefore, for all PAHs, the respective PR values for the pristine membrane were lower than the values obtained for the MMMs, suggesting that the incorporation of PABA-MCM-41 into the PES facilitated the permeation of PAHs through the PES MMMs. However, the achieved values for the PES MMM with 7.5 wt.% were not so much higher than those found for the pristine membrane, as well as for the other MMMs.
In the RR values, which have an inverse relationship to the PR values, it was noted that the results found for the pristine PES membrane were higher than the results of the MMMs, except for MMM with 7.5 wt.%, whose RR values are close to the pristine membrane. On the other hand, the incorporation of PABA-MCM-41 within the MMMs caused a reduction of the removal percentage. From our knowledge, we found only one published article reporting the PAHs remediation using MMMs [14]. This mentioned work was recently published by our research group, in which the removal percentages obtained for B 190%, for the same PAHs, by the pristine PES membrane was higher than that observed for PSf-Ac-and PES-based MMMs. Figure 5(a-c) shows the correlation between the PR and T g values of PES-based MMMs. It is possible to observe an inverse relation between these values, except for MMM with 2.5 wt.%, in which the increase of PR values occurred because of the decrease in T g values. This comportment can be attributed to the increase in the mobility of the PES chains, promoting more easily the permeation process of PAHs. However, this phenomenon was less meaningful for the MMM with 7.5 wt.% of mesoparticles.  On the contrary (Fig. 5d-f), a direct relationship between the PR and WVT values. On the one hand, the decrease in Tg affected in the growth in the PR values. This decrease also boosted an increase in the WVT values. This comportment can be attributed to the growth in the number of preferred paths in the PES-based MMMs with the filling of mesoparticles, in addition to the mobility of the PES chains.
Similarly, to PR values, the RR values were also influenced by the T g values, as shown in Fig. 5(g-i). A direct relationship was observed between the RR and T g values, less for the MMM with 2.5 wt.%. In this way the decrease of the RR occurred due to the decay of Tg values. However, the decrease in T g was less evident for MMM with 7.5 wt.% of PABA-MCM-41. As a direct consequence, the decrease in retention rate (RR) was directly influenced by the increase of the permeation rate (PR) of PAHs mixture. In this way (Fig. 6(a-c)), it is possible to observe an inverse correlation between the RR and WVT values for PES MMMs. Figure 6(d-f) and 6(g-i) present the correlations of the removal percentage of PAHs mixture with the Tg and WVT values, respectively. From Fig. 6(d-f), for some cases, that the removal percentage values do not have a defined relationship with the T g values. It was expected larger removal percentages at higher Tg values, as well as lower removal percentage values at lower Tg values. This behavior is due, respectively, to a greater or lesser ease of permeation of PAHs through the polymer chains of the MMMs. In this sense, this tendency was noted in the B[b]F and B[k]F removal, in the   These results show, for these MMMs, that the polymer matrix has more contribution in the adsorption of PAHs than the mesoporous array. But, for the other PES-based MMMs, we can observe an inverse relation between the removal percentage and Tg values, suggesting that the mesoporous array has a greater contribution in the adsorption of these PAHs than the PES matrix. Finally, from the analysis of Fig. 6

Conclusion
In summary, we report the preparation of a novel mixed matrix membrane prepared with PES polymer and PABA-MCM-41 modified mesoporous array, which was applied in the PAHs removal. The hydrothermal/co-condensation method was useful, low-cost, and sustainable in the preparation of eco-friendly PABAMCM-41 matrix with an architectured hexagonal arrangement and interesting mesoparticles features.
In the same way as the casting approach was useful in the formulation of the PES-based mixed matrix membranes, which exhibited a remarkable and efficient dispersion of the Funding The authors wish to thank Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) and Centro de Desenvolvimento de Materiais Funcionais for providing the essential financial support.

Declarations
Consent for Publication The Authors transfer to Springer the publication rights.