4.2.1 Effect of pH
Acidity of the solution affects the solubility (ionization) of the metal ions and concentration of the counter ions of the surface groups (in this case H+) [119]. These significantly influence the adsorption process and therefore pH should be optimized to ensure the best adsorption efficiency.
Figure 11 shows the effects of pH on the adsorption of Co(II) by EDTA-functionalized adsorbents. It can be seen that adsorption efficiency increased with the increasing pH for all the adsorbents. The similar trend was observed for almost all the EDTA/DTPA-functionalized
materials presented in Tables 6 and 7. The limiting behavior at low pH can be attributed to the competition between protons and metal ions for the available adsorption sites [115]. Zeta-potential measurements showed that the surface charge of the EDTA/DTPA-functionalized silica gels increased as the pH decreased (Figure 2 in Paper I) suggesting repulsive forces between surface and metal cations in acidic media.
1 2 3 4 5 6
0 20 40 60 80
100 EDTA-silica gel
EDTA-Chi:TEOS 2:60 EDTA-Chi:TEOS 2:30 EDTA-Chi:TEOS 2:15 EDTA-chitosan
% Co(II) adsorbed
pH
Figure 11. Effect of pH on adsorption of Co(II) by EDTA-functionalized adsorbents.
Experiments repeated using the initial Co(II) concentration of 1.3 mM. Dose: 2 g/L.
Comparison between different materials suggests that higher chitosan content increased the adsorption efficiency of the material at low pH (see also Figure 3b in Paper III). For example, at pH 2 81% of maximum capacity was obtained for EDTA-silica gel, 87% for EDTA-Chi:TEOS 2:15, and 95% for EDTA-chitosan. It is possible that the chitosan matrix, which is more electronegative (electronegativity of carbon: 2.5) than the silica gel matrix (electronegativity of silicon: 1.7) increased the acidity of the EDTA carboxyl groups making them more available for metal adsorption at low pH range. Moreover, the chelating ability of immobilized EDTA-groups at a low pH has also been observed elsewhere and assigned to the inductive effects that decrease pKa values of EDTA carboxyl groups [122].
In Paper II (Figure 2b), it was observed that EDTA/DTPA-functionalized chitosans showed a clear adsorption maximum for Co(II) at pH range from 2 to 2.5 at low metal concentrations (17 µM). This was attributed to the crosslinking effect between carboxyl groups of
chelating agents and the surface. At higher concentrations fast metal binding occurred before crosslinking and adsorption capacity remained similar after pH 2. On the other hand, a similar effect was not seen in modified silica gels possibly due to their rigid structure as well as a lower amount of primary amino groups preventing corsslinking (see FTIR spectra in Figure 8).
The type of the target metal significantly influenced the pH curves (Papers I-III). Ni(II) started to adsorb at a lower pH range than Co(II) and at pH as low as 1 around 90% Ni(II) removal was obtained for EDTA-chitosan from a 100 mg/L solution (Figure 2 in Paper II). This indicates that EDTA/DTPA-functionalized adsorbents could be used to separate Ni(II) from Co(II) at a low pH, which could be applied in hydrometallurgy, for example [142]. The Cd(II) adsorption had a pH behavior similar to Co(II) and Pb(II) to Ni(II) (Figure 3a in Paper III).
Better adsorption efficiencies of Ni(II) and Pb(II) over Co(II) and Cd(II) can be explained by the higher stabilities of their EDTA- and DTPA-chelates forming at low pH (see Table 4 and Appendix I).
4.2.2 Effect of contact time
Figure 12 shows the effect of contact time on the adsorption of Co(II) by EDTA/DTPA-functionalized adsorbents. At the initial adsorption stage, the adsorbent surface contained a lot of available active sites for metal binding and fast adsorption took place. After this adsorption slowed down due to the decrease of concentration of metals in solutions phase as well as a possible location of active sites in the positions that were not easily available (for example inside the pores).
For all the EDTA-functionalized adsorbents a contact time of 6 h was sufficient to attain equilibrium at high metal concentrations (1.3 mM). More differences were seen at lower metal concentrations (Figure 12b), for which EDTA-chitosan showed the fastest adsorption. This can be attributed to the highest surface coverage of EDTA on the chitosan surface creating active sites that were easily available for metal binding. In the case of EDTA-Chi-TEOS 2:60 and EDTA-silica gel the smaller size of the former (see Table 8 and Figure 7) led to faster adsorption, which was also evident for the modified silica gels of different particle sizes (Figure 7 in Paper I).
0 400 800 1200 1600 0.0
0.2 0.4 0.6 0.8
qt (mmol/g)
EDTA-silica gel (40-63 µm)
EDTA-Chi:TEOS 2:60 EDTA-Chi:TEOS 2:30 EDTA-Chi:TEOS 2:15 EDTA-chitosan
t (min) qt (mmol/g)
t (min) (a)
0 400 800 1200 1600
0.00 0.04 0.08 0.12
0.16 (b)
EDTA-silica gel (40-63 µm)
DTPA-silica gel (40-63 µm)
EDTA-Chi:TEOS 2:60 EDTA-chitosan DTPA-chitosan
Figure 12. Effect of contact time on adsorption of Co(II) by EDTA/DTPA-functionalized adsorbents. Results collected from Papers I-III and V. Dose: 2 g/L, pH: 2 for modified chitosans and pH: 3 for other adsorbents, Co(II) concentration: (a) 1.3 mM and (b) 0.3 mM.
For modified silica gels, the effect of contact time was not significant when EDTA- and functionalized materials were compared (Figure 12b). Differences between EDTA- and chitosan were clear (Figure 12b, Figure 4 in Paper II), however, and slower adsorption on DTPA-chitosan was attributed to the crosslinking caused by bulky DTPA-groups [111] and lower surface coverage of DTPA.
4.2.3 Effect of initial metal concentration
Adsorption isotherms were measured by varying only the initial metal concentrations.
Comparisons of different EDTA-functionalized materials are shown in Figure 13. The highest adsorption capacity was obtained for EDTA-chitosan, which was expected due to its highest ligand loading. Furthermore, the adsorption capacity for hybrid adsorbents increased as the chitosan content increased. The adsorption efficiencies of EDTA-silica gel and EDTA-Chi:TEOS 2:60 were, however, similar.
0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0
0.2 0.4 0.6 0.8 1.0
1.2 EDTA-silica gel
EDTA-Chi:TEOS 2:60 EDTA-Chi:TEOS 2:30 EDTA-Chi:TEOS 2:15 EDTA-chitosan qe (mmol/g)
Ce (mmol/L)
Figure 13. Effect of initial metal concentration on the adsorption of Co(II) by EDTA-functionalized adsorbents. Dose: 2 g/L, pH: 2 for EDTA-chitosan and pH: 3 for other adsorbents.
Results collected from Papers I-III
Table 10 further summarizes the adsorption efficiencies. It is interesting to note that even if the modified chitosans showed the highest adsorption efficiency at high concentrations they gave lower efficiency at concentrations below 0.17 mM (10 mg/L). A similar effect was seen in the pH experiments (see section 4.2.1) and can be attributed to the enhanced interactions of carboxylic groups and chitosan surface at low metal concentrations (crosslinking effect).
Table 10. Comparison of adsorption efficiencies of EDTA/DTPA-functionalized adsorbents.
Co(II) conc. 0.17 mM Co(II) conc. 0.85 mM Co(II) conc. 1.7 mM Adsorbent pH qm (mmol/g)a % of Co(II)
adsorbed
% of Co(II) adsorbed
% of Co(II) adsorbed
EDTA-silica gel 3 0.34 97.4 58.7 35.5
DTPA-silica gel 3 0.27 99.0 62.7 31.6
EDTA-chitosan 2 1.07 98.0 96.6 93.8b
DTPA-chitosan 2 0.83 95.5 93.3 85.9
EDTA-Chi:TEOS
2:60 3 0.25 98.6 64.3 30.2
EDTA-Chi:TEOS
2:30 3 0.42 98.2 83.9 55.6
EDTA-Chi:TEOS
2:15 3 0.63 98.4 93.5 67.7
aData collected from Papers I, II, and III
b99.2% at pH 3
Comparison of Table 10 to Tables 6 and 7 shows that our EDTA- and DTPA-silica gels had considerably higher adsorption efficiencies than those obtained by Shiraishi et al. [110] despite their equal or higher coverage of functional groups. This discrepancy can be explained at least partly by the different experimental conditions. Shiraishi et al. used only 1 hour equilibrium time and pH 1 where they also observed the highest adsorption performance. This was not the case in this study and therefore, it is possible that the materials prepared by Shiraishi et al. differed from those synthesized in our work.
EDTA- and DTPA-chitosan were also synthesized earlier by Inoue et al. [113]. They obtained somewhat higher adsorption capacities compared to this work, which was attributed to the differences between the starting materials (Paper II). The adsorption performances of EDTA-chitosan silica hybrid adsorbents were comparable to many of those presented in Table 6.
Especially, EDTA-Chi:TEOS 2:15 clearly showed a higher adsorption capacity compared to conventionally functionalized silica gel. This novel hybrid adsorbent, having a rigid structure (see section 4.3 and Figure S2 in Paper III), seemed to combine the beneficial properties of both silica gel and chitosan.
4.2.4 Adsorption mechanism
Papers I and II present the suggested adsorption mechanisms for EDTA-silica gel and -chitosan.
Furthermore, FTIR-measurements confirmed that metals were bound on the surface by chelation
(Figures 9 and 10). In the case of EDTA- and DTPA-functionalized silica gels (pH 3), the release of protons during the adsorption was approximately one to every adsorbed metal ion, which for Co(II) adsorption can be explained by the following reaction equation:
SiO2-NH-HEDTA2- + Co2+↔ SiO2-NH-Co(II)EDTA- + H+ (47)
The above mechanism is also supported by the zeta-potential measurements, which showed that the surface carried a negative charge after metal binding (Figure 2 in Paper I). For modified chitosans the solution pH was not significantly affected by metal adsorption. The mechanism derived from the speciation calculations, however, showed that some protons should have appeared in the solution due to the adsorption (see Section 3.3 in Paper II). This discrepancy was explained by the binding of released protons by free amino groups on the chitosan surface [145].
In contrast to these results, EDTA-functionalized chitosan-silica hybrid materials showed a release of two protons for every adsorbed metal ion at a pH range from 3 to 5. Therefore, the following reaction is suggested:
SiO2-NH-H2EDTA- + Co2+↔ SiO2-NH-Co(II)EDTA- + 2H+ (48)
Unlike for modified chitosans, released protons were not able to bind on the surface, because the hybrid chitosan-silica network (Figure 6) did not contain any free amino groups after EDTA-immobilization (see Figure S1 in Paper III and ref. [146]). On the whole, however, the surface reactions are better considered as chelation rather than ion-exchange due to the uncertainty of the true amount of protons released for every metal ion absorbed. In addition, for example at neutral conditions, all the carboxyl groups of the immobilized chelating agents should be in their ionic form when ion-exchange is not longer possible.
Finally, it is important to note that speciation calculations based on the aqueous species may not be directly applicable for the surface bound chelates [5]. Therefore, the mechanisms presented in this work are only suggestions based on the best available information.
4.3 Stability and regenerability of the EDTA/DTPA-functionalized adsorbents