Characterization
We utilized FT-IR and solid-state 13C NMR to confirm the functional chemical groups in Cr(VI)IMS. Figure 2a is the IR spectra of Cr(VI)IMS. Curve 1 and 2 represented the IR absorption signals of MCM-41 and Cr(VI)IMS, respectively. Three peaks at 1078 cm-1, 941 cm-1 and 795 cm-1 indicated the silica matrix. The bending vibration of O-H of silica materials was displayed at 1632 cm-1(Wang et al. 2014). The broad band at 3454 cm-1 of curve 1 might attributed to the existence of Si-OH. The weak adsorption peak of curve 2 at 1577 cm-1 was ascribed from the vibration of imidazole C-N bond(Kryszak et al. 2017). The solid-state 13C NMR spectra of Cr(VI)IMS were exhibited in Fig. 2b. The peaks at 8 ppm, 24 ppm and 50 ppm of Cr(VI)IMS corresponded to C1, C2 and C3, while signals at 121 ppm and 134 ppm were assigned to the C atoms in imidazole ring(Li et al. 2007). The FT-IR and NMR spectra demonstrated that the functional monomers have poly-condensed into Cr(VI)IMS.
Small angle XRD patterns of MCM-41 and Cr(VI)IMS were depicted in Fig. 2c. Two materials showed three diffraction peaks of (100) (2θ = 2.0o), (110) (2θ = 3.8o) and (200) (2θ = 4.4o) lattice plane. This results indicated that Cr(VI)IMS still possessed the typical mesoporous structure of mesoporous silica MCM-41.
The mesoporous features of Cr(VI)IMS were investigated with N2 adsorption-desorption experiment. The isotherm of Cr(VI)IMS was of the type IV(Fig. 2d), suggesting that the pores of Cr(VI)IMS were in the mesoporous range. The specific surface area (1054.51 m2 g-1) and pore volume (1.16 m3 g-1) were calculated with Bnmauer-Enunett-Teller (BET) method. The BJH-determined average pore size distribution was 3.45 nm, with pore sizes largely in 2.44–3.63 nm range. The test results corresponded to the characteristics of mesoporous silica MCM-41.
The morphology and microstructure of Cr(VI)IMS was further observed by SEM and TEM. Cr(VI)IMS mainly existed as dispersed microspheres with particle diameter of 100–400 nm in SEM images (Fig. 3a). TEM images (Fig. 3b) demonstrated the ordered 2D hexagonal symmetrical structure of mesoporous channels with pore diameter of 2–3 nm inside Cr(VI)IMS, which basely corresponded with the result of N2 adsorption-desorption experiment.
Effect of pH on adsorption capacity of Cr(VI)IMS
The acidity of solution poses obvious effects to the absorption performance of Cr(VI)IMS. As displayed in Fig. 4a, at low pH, Cr(VI)IMS showed high absorption ability. Cr(VI)IMS achieved the largest binding capacity (45.6 mg g-1). The protonation of imidazole groups of Cr(VI)IMS were positively charged at low pH. With the increase of pH values, deprotonation increased simultaneously, which can lead to intense decrease of adsorption capacity(Amara and Kerdjoudj 2003). While the pH value of the solution was higher than 7, OH- was able to neutralize the positive charge of adsorbents, which caused the binding capacities of Cr(VI)IMS obviously decreased. We used imprint factor(IF) value to estimate the specific adsorption ability of Cr(VI)IMS. The calculation formula of IF value is IF = QI / QN, where QI and QN represent adsorption capacity of Cr(VI)IMS and NIMS, respectively. Above all, the adsorption abilities of Cr(VI)IMS was mainly determined by the extent of protonation of imidazole groups. The imprinted sites of Cr(VI)IMS exhibited higher affinity to Cr(VI) ions at higher pH, while Cr(VI)IMS showed higher adsorption capacity to Cr(VI) at acidic solution.
Adsorption isotherm study
The adsorption isotherm experiments were carried out at the optimized pH condition (pH = 3), as the initial concentration of Cr(VI) ions increased, the adsorbing capacity increased gradually(Fig. 4b). When the initial concentration of Cr(VI) ions was 200 mg L-1, Cr(VI)IMS reached saturated adsorption (45.6 mg g-1), which was much higher than that of NIMS. The IF value was 2.36, which indicated that imprinted functionalized mesoporous exhibited high affinity to given targets.
We used Scatchard model to analyze the results of adsorption isotherm experiments(Gao et al. 2011). The Scatchard equation is Qe/Ce=(Qmax-Qe)/KD, where Ce, Qmax and Qe represent equilibrium concentration of Cr(VI) ions, apparent maximum adsorption capacity and equilibrium adsorption capacity. KD is the equilibrium dissociation constant. The results of the linear regression Scatchard scatter plots of Cr(VI)IMS displayed two separate straight lines with different slopes, suggesting two types of recognition sites (specific binding and nonspecific binding) existed in the Cr(VI)IMS(Li et al. 2014). On the other hand, the scatter plots could be modelled as one straight line, indicating that there was one kind of binding sites NIMS. (Fig. 5 and Table 1)
Table 1
Results of Scatchard analysis
Adsorbent
|
linear equation
|
apparent maximum adsorption capacity Qmax(mg g− 1)
|
equilibrium dissociation constant KD(mg L− 1)
|
R2
|
Cr(VI)IMS
|
1:Qe/Ce=-0.05072Qe + 1.2029
|
23.72
|
19.72
|
0.9908
|
2:Qe/Ce=-0.00479Qe + 0.5100
|
106.47
|
208.77
|
0.9907
|
NIMS
|
3:Qe/Ce=-0.00548Qe + 0.2132
|
38.91
|
182.48
|
0.9909
|
The Langmuir adsorption model was also used to assess the adsorption thermodynamic properties(Fig. 6). The expression is Ce /Qe = Ce /Qmax+1/ (Qmaxb), where Ce, Qmax and Qe represent equilibrium concentration of Cr(VI) ions, the maximum amount of adsorption and equilibrium adsorption capacity. And b is the constant in the Langmuir adsorption model. The adsorption behavior basically conformed to the Langmuir adsorption model (correlation coefficient R2 = 0.9878), indicating that surface adsorption was rather uniform. The maximum amount of adsorption Qmax and the experimental amount QExp are 55.9 mg g-1 and 45.6 mg g-1, respectively.
Adsorption kinetic
We studied the adsorption kinetic to Cr(VI) ions of Cr(VI)IMS at the concentration of 50 mg g-1 and 200 mg g-1(Fig. 7). At low concentration, it took Cr(VI)IMS 7 min to achieve equilibrium. For 200 mg g-1, the adsorption equilibrium time of Cr(VI)IMS was 10 min. The results suggested that higher concentration led to longer absorption time for more Cr(VI) ions to enter the imprinted cavities. The recognition sites in nanoscale pore walls of large surface area accelerate the mass transportation. Compared with the ion imprinted polymers of classic structure, the functionalized mesoporous micromorphology remarkably shorten the adsorption time of Cr(VI)IMS to Cr(VI) ions(Bayramoglu and Arica 2011).
Kinetic studies were explored by two equations including the pseudo first order and the pseudo second order model. The fitting plots of kinetic equations are shown in Fig. 8a and 8b. The pseudo first order kinetic equation is qt = qe [1-exp(-k1t)] and the pseudo second order kinetic equation is qt = k2qe2t/(1 + k2qet), where t, qt and qe represents adsorption time, the amount of adsorption and equilibrium adsorption capacity. k1 and k2 are adsorption rate constants of the pseudo first order and the pseudo second kinetic, respectively. The related parameters are presented in Table 2 and Table 3. It is obvious that the pseudo-first-order model was better than the pseudo-second-order model to describe adsorption kinetics as shown by the responding regression coefficient (R2).
Table 2
Calculated kinetic parameters of pseudo-first orders
Adsorbent
|
concentration (mg L− 1)
|
R2
|
k1(min− 1)
|
qe (mg g− 1)
|
Cr(VI)IMS
|
50
|
0.9934
|
0.3784
|
15.74
|
Cr(VI)IMS
|
200
|
0.9888
|
0.2970
|
46.52
|
Table 3
Calculated kinetic parameters of pseudo-second orders
Adsorbent
|
concentration (mg L− 1)
|
R2
|
k2(g mg− 1 min− 1)
|
qe (mg g− 1)
|
Cr(VI)IMS
|
50
|
0.9668
|
0.0274
|
17.79
|
Cr(VI)IMS
|
200
|
0.9587
|
0.0065
|
54.02
|
Selective experiment
Figure 9 was the results of selective experiments. The functional chemical groups of Cr(VI)IMS were organized during the imprinting process and fixed in the imprinted silica. Because of the existence of complementary imprinted cavities for Cr(VI) ions, the Cr(VI)IMS exhibited obviously selective binding ability to Cr(VI) ions. H2PO4- and NO3- are common anions. And they were adsorbed by positive charged Cr(VI)IMS and NIMS via electric attraction. Distribution coefficient K = Q/Ce, where Q and Ce were the adsorption capacity of adsorbents and the concentration of residual ions. We used selective coefficients RSF = KI/KN to estimate the selectivity of three Cr(VI)IMS, where KI and KN represented distribution coefficient of Cr(VI)IMS and NIMS. The RSF values of Cr(VI)IMS (2.56) were much higher than other two anions (Table 4), which could signified H2PO4- and NO3- were hard to be captured by the recognition sites due to their differences with Cr(VI) ions. On the other hand, functional monomers were distributed randomly in the materials and showed nonspecific adsorption to anions.
Table 4
Results of selective adsorption experiment of Cr(VI)IMS and NIMS
Adsorbent
|
Ions
|
QI(mmol g− 1)
|
QN(mmol g− 1)
|
KI(L g− 1)
|
KN (L g− 1)
|
RSF
|
|
Cr2O72−
|
0.294
|
0.140
|
416.4
|
162.8
|
2.56
|
Cr(VI)IMS
|
H2PO4−
|
0.168
|
0.132
|
201.9
|
152.1
|
1.33
|
|
NO3−
|
0.134
|
0.113
|
154.7
|
127.4
|
1.21
|
Reusability and reproducibility
Reusability plays a significant role in the application of adsorbents. To estimate the reusability of Cr(VI)IMS, adsorption-regeneration cycling experiments were conducted on the same batch of Cr(VI)IMS. After 8 cycles, the adsorption efficiency of Cr(VI)IMS still reached 92.5%, showing their excellent stability and reusability(Fig. 10a). Reproducibility is another important property of adsorbent. Five batches of Cr(VI)IMS exhibited no obvious difference in adsorption capacities(Fig. 10b).
Application in real water samples
The concentration of Cr(VI) of the filtered waste water was 22.1 mg L-1. The water samples of the same volume were treated by different amount of adsorbent. The adsorption performance to actual waste water was shown in Table 5. Lower residual concentration can be obtained, treated by larger amount of adsorbent. Using 6.0 g L-1 of Cr(VI)IMS, the residual concentrations of Cr(VI) ions was 0.2 mg L-1, which was lower than discharge standard of China (GB8978-1996).
Table 5
adsorption performance of different adsorbents to actual waste water
Adsorbent
|
Initial concentration
(mg L− 1)
|
Amount of adsorbent(g L− 1)
|
Residual concentration
(mg L− 1)
|
Cr(VI)IMS
|
22.1
|
2.0
|
5.1
|
Cr(VI)IMS
|
22.1
|
3.0
|
2.9
|
Cr(VI)IMS
|
22.1
|
4.0
|
1.5
|
Cr(VI)IMS
|
22.1
|
5.0
|
0.6
|
Cr(VI)IMS
|
22.1
|
6.0
|
0.2
|
Comparison of adsorption performance of other adsorbents
To study removal of Cr(VI) ions by adsorption method in the water, several adsorbents have been utilized and their performance were summarized in Table 6. Fe3O4 50%-PANI@GO hybrid composite, imidazole functionalized mesoporous SBA-15 and ion-exchange resin are three kinds of nonspecific anion adsorbents. The adsorption capacities of Cr(VI) ions are higher than that of Cr(VI)IMS. However, the equilibrium time is much longer and they have not shown high selectivity to Cr(VI). After 5 cycles regeneration, the adsorption capacity of Fe3O4 50%-PANI@GO hybrid composite has decreased remarkably while Cr(VI)IMS with relatively robust structure has shown reliable reusability. Cr(VI) imprinted polymers and magnetic Cr(VI) imprinted nano composites mentioned in Table 6 have been applied imprinted technology, so they have performed well in selective adsorption but adsorption time is longer than that of Cr(VI)IMS. According to the comparison of adsorption capacities, adsorption time and selectivity, it can be concluded that ion imprinting technology enhance the selectivity of adsorbents and Cr(VI)IMS prepared by co-condensation method exhibited the advantages of fast adsorption rate, high binding capacity and specific adsorption.
Table 6
Comparison of the reported adsorbents for hexavalent chromium
Adsorbents
|
Functional groups
|
Synthesis method
|
Adsorption capacity (equilibrium time)
|
Selectivity
|
References
|
Fe3O4 50%-PANI@GO hybrid composite
|
aniline groups
|
(multi-step)
|
98.5 mg g− 1 (50 min)
|
Not obvious
|
(Chinnathambi and Alahmadi 2021)
|
Imidazole functionalized mesoporous SBA-15
|
Imidazole groups
|
Chemical modified (multi-step)
|
113 mg g− 1 (60 min)
|
Not mentioned
|
(Li et al. 2007)
|
Ion-exchange resin
|
Tertiary amino
|
Chemical modified (one-step)
|
152.52 mg g− 1(60 min) ; 120.48 mg g− 1(60 min); 156.25 mg g− 1 (60 min)
|
Not mentioned
|
(Shi et al. 2009)
|
Cr(VI) imprinted polymers
|
Quaternizated pyridine
|
Free radical polymerization, quaterisation (one-step)
|
45.23 mg g− 1 (90 min)
|
Cr(VI) ions
|
(Pakade et al. 2011)
|
Magnetic Cr(VI) imprinted nano composites
|
Amino and imidazole
|
Free radical polymerization and sol-gel method (one-step)
|
2.49 mg g− 1 (30 min)
|
Cr(VI) ions
|
(Qi et al. 2017)
|
Cr(VI)IMS
|
Imidazole groups
|
Sol-gel method (one-step)
|
45.6 mg g− 1 (10 min)
|
Cr(VI) ions
|
This work
|