3.1. Catalysts Characterization
Diffraction patterns from three iron-doped carbon xerogel catalysts (Fe-CX) are presented in Figure 2
. In each case, three phases were identified; graphite (JCPDS 75-1621), iron ferrite (JCPDS 06-0696), and a cohenite (Fe3
C) phase (JCPDS 34-001). At least one other unidentified phase is present (unidentified peak located at 51.3°). Diffraction from the graphite 002 planes occurred at 26.23° in 2 θ.
This peak was used to scale the intensities of the three datasets for comparative purposes. Elemental iron was present in the xerogel catalyst in two phases, a cubic iron ferrite phase and an orthorhombic cohenite iron carbide phase indicating reduction of Fe3+
during the pyrolysis process. The iron ferrite phase was identified by diffraction from the 110 and 200 sets of planes (diffracted peaks at 44.67° and 65.02° in 2θ
, respectively). These peaks are denoted with a dotted line in Figure 2
. Notice, relatively weak diffraction from the graphite 101 planes overlap that from the iron ferrite 110 planes (44.36°). The cohenite iron carbide phase is identified by the 102, 211, 112, and 221 planes located at 43.74°, 42.86°, 45.85°, and 49.10°, respectively. Iron-doped carbon xerogel catalyst Fe-CX(1) represents a base line diffraction pattern where the sample was washed with ethanol and underwent ion exchange for the addition of Fe3+
before pyrolysis. Catalyst Fe-CX(2) underwent a 72 h washing with water before ion exchange while, Fe-CX(4), underwent a 72 h washing after ion exchange but before pyrolysis, Table 1
. The ordering of the washings proved to be very significant and demonstrated large variations in the iron content of the resulting catalyst. The XRD data show a significant increase in the iron ferrite content in the diffractogram for Fe-CX(2) with an accompanying decrease of the iron carbide content. The diffractogram for Fe-CX(4) indicates a significant decrease of the iron ferrite component (when looking at the diffractogram, it should be remembered that diffraction from the graphite 101 planes overlaps that from the iron ferrite 110 planes). In this sample, the iron carbide content is also substantially reduced.
FT-IR spectra of CX and Fe-CX samples are presented in Figure 3
. For better visualization of the structural modifications that took place after iron doping, infrared bands were indexed with respect to their position in carbon xerogel. The recorded spectral range was plotted into two figures for a better analysis of bands evolution after metal doping and pyrolysis steps.
In the 400–2000 cm−1
range, Figure 3
a, absorption bands appear at 670, 702, 795, 886, 1059, 1375, 1443, 1473, 1580, 1740 cm−1
for CX sample. Most of these bands are narrow and reveal an organized structure inside the carbon xerogel CX. In this range, vibrations related to carbon and functional units are expected. Starting with the less energetic vibrations in 650–750 cm−1
range, two narrow signals can be noticed, located at ~670 and ~702 cm−1
. On the right side of these bands, a broad and less intense envelope appeared. This more energetic spectral domain is composed of two signals, one located at 782 cm−1
and the other one at ~818 cm−1
. The [650–850] cm−1
spectral domain was attributed to the aromatic ring in particular to C–H bond vibrations [5
]. More energetically Lorentzian band at 886 cm−1
was ascribed to stretching vibration of C−H bonds, which seems to have similar length and surroundings within the rigid structure of carbon xerogel [31
Two intense vibrations located at 1375 and 1443 cm−1
and a shoulder at 1473 cm−1
connected to the last vibration belong all to bonds linked to aromatic rings [5
]. After iron doping, all the above-mentioned bands disappeared under a broad infrared envelope. An interesting behavior was observed for the band at 1730 cm−1
related to carboxyl in esters group; more precisely, to the stretching vibration of υ(C=O) [44
] in [COO]-K as obtained during synthesis process that cannot be found in the iron-doped samples.
The 2000–4000 cm−1
spectral range presented in Figure 3
b is clearly dominated, in the carbon xerogel sample, by a strong band located at 3198 cm−1
, with two shoulders at around 3000 and 3452 cm−1
. After iron doping, the last mentioned band became the most prominent in the spectra while the first two are hidden under a broad envelope. The band at 3450 cm−1
might be assigned to residual –OH groups from the gel precursors and adsorbed water [31
]. The bands at 2852 and 3002 cm−1
were associated with asymmetric and symmetric vibrations of aliphatic and aromatic structures [46
]. Infrared bands, which build up the broad band with a metacenter at 2294 cm−1
, were assigned to −C≡C− triple bond [5
SEM-EDX analyses on carbon xerogel and iron-doped carbon xerogel samples revealed the strong connection between the synthesis procedure, specifically washing steps (Table 1
), and iron content of the final sample. For CX sample, a porous carbon matrix (Figure 4
a) with elements originating from the inorganic and organic precursors used for synthesis were identified (Figure 4
b and Table 2
). As potassium was replaced with iron during the ion exchange step, several differences were observed. Samples Fe-CX(1) and Fe-CX(2) have a similar average iron content as no water washing step was included after the ion exchange process, 7.0% and 7.3%, respectively (Table 2
). This difference might be attributed to an improved ion exchange process, for Fe-CX(2), as it took place on a water prewashed organic gel (Table 1
). For Fe-CX(3) sample, which synthesis included two water washing steps, before and after the ion exchange process, average surface iron content dropped to less than half (Table 2
). In the case of Fe-CX(4) and Fe-CX(4.1) samples, where water washing was performed only after the ion exchange process, iron content dropped significantly (Table 2
). SEM-EDX iron maps presented in Figure 4
for Fe-CX(2) and Fe-CX(4.1) samples show iron distribution on the surface and the decreased content in Fe-CX(4.1) correlated with the synthesis steps discussed above.
TEM images presented in Figure 5
for Fe-CX(2) and Fe-CX(4) reveal how iron is imbedded in the carbon (graphite) matrix as well as the decreased amount in Fe-CX(4) sample.
3.2. Dyes Oxidation Tests
Iron-doped carbon xerogel samples, prepared as described above, were tested in heterogeneous Fenton and CWAO processes. Experiments were conducted in the same conditions (see Section 2.4
) for BG, CV, and MG dyes. The obtained results are presented in Figure 6
, Figure 7
and Figure 8
for heterogeneous Fenton and Figure 9
, Figure 10
and Figure 11
for CWAO (for UV-Vis spectra, *indicates a dilution coefficient of 10). E and ETOC
(%) values are very well correlated with the iron content, decreasing with the amount of iron in the catalysts. Dye removal efficiencies up to 99% were obtained for all dyes and for both oxidation processes. In terms of total organic carbon efficiencies, the highest values were in the 50%–65% range for heterogeneous Fenton and in the 66%–92% range for CWAO. The CWAO results are very promising as the reaction took place under atmospheric pressure and at only 25 °C. The general trend observed was: Highest efficiencies for Fe-CX(1), Fe-CX(2), and Fe-CX(3) and between the three of them, for Fe-CX(2), the highest efficiency was recorded in most of the cases, in correlation with iron content (Table 2
). For heterogeneous Fenton process, both E and ETOC
% efficiency values decrease in the following order: MG > BG > CV for Fe-CX(2) catalyst. In the case of CWAO, E % values are very close to each other for the three dyes, 98.3%–98.8%, while ETOC
% values decrease in the following order: BG > MG > CV.
In terms of oxidation process evolution, UV-Vis spectra indicated that chemical composition of the reaction mass changed dramatically as absorption maxima shifts from the visible region towards UV region as non-colored reaction intermediates start to form, especially in the case of heterogeneous Fenton. As maximum absorption peaks at 625, 584, and 631 nm for BG, CV, and MG, respectively, decreased drastically by the end of the oxidation process, a new absorption maximum was recorded in 200–250 nm region. In the case of MG, a second, higher absorption peak was recorded at about 375 nm. For the CWAO process, organic content considerably dropped, as shown in both UV and visible regions, with best results for BG and MG dyes.