3.2. FTIR Spectra
By Fourier transform infrared (FTIR) spectroscopy, functional groups present on the adsorbent surfaces were identified. The FTIR spectra of the 12 food waste materials in the range of 5000–600 cm−1
are shown in Figure 2
The broad band in the range of 3700 and 3000 cm−1
includes many vibrations modes corresponding to −OH of alcohols [44
], phenols and carboxylic acids [36
]. This region is greater in watermelon peel and tomato peel spectra, whereas it appears reduced in eggplant peel and carob peel spectra. The two sharp bands at 2925 cm−1
and 2855 cm−1
belong to the C−H bonds of methyl and methylene groups of lipids [44
]. These bands are more intense in banana peel, apple peel, coffee waste and decaf coffee waste spectra. The band at 1742 cm−1
corresponds to the C=O vibration in the carbonyl group of −COOH. The bands in the range of 1600−1400 cm−1
can be ascribed to C=C vibration of lipids, fatty acids and lignin moieties [36
]. Figure 2
shows that the bands from 1742 cm−1
to 1400 cm−1
are less intense in potato peel, carob peel and eggplant peel spectra. The region at 1400−900 cm−1
shows several types of vibrations including C−H, C−O−C, C−N and P−O of polysaccharides [44
]. This region is greater in watermelon peel, tomato peel, lemon peel and orange peel spectra.
By the recorded FTIR spectra, similarities and differences of the chemical structures of the adsorbent surfaces were identified. Figure 2
reveals that each food waste material has different functional groups which may act as selective active sites for metal and metalloid ions coordination.
The spectral data obtained by FTIR spectroscopy were elaborated by performing multivariate statistical computations. Principal component analysis (PCA) of the spectral data allowed to group the food waste materials according to functional groups present on their adsorbent surfaces.
shows that the food waste adsorbents were grouped in five main clusters (marked in different colors) on the first two principal components (PC1 and PC2) of the score plot, which explain 75.7% of the total explored variance. A first cluster, composed by watermelon peel and tomato peel (red color) is placed on the bottom left part of the score plot’s center. The adsorbents of this cluster showed the highest amounts of −OH of alcohols, phenols and carboxylic acids; C−H, C−O−C, C−N and P−O of polysaccharides; and C=C of lipids and lignin moieties. Lemon peel and orange peel (brown color), which are placed on the upper left compared to the first cluster (red color), have a lower amount of these functional groups. Potato peel is away from lemon peel and orange peel mainly because of the lower amount of C−H, C−O−C, C−N, and P−O of polysaccharides and C=C of lipids and lignin moieties. The other two clusters are on the right of the score plot. The cluster above the score plot’ center (blue color) is composed by the adsorbents which showed the highest amount of C−H of methyl and methylene groups and lower amounts of the other functional groups: apple peel, banana peel, decaf coffee waste and coffee waste. The other one (green color), below the score plot’ center, is formed by eggplant peel, carob peel and grape waste which revealed the lowest amount of functional groups on their adsorbent surfaces.
3.3. Removal Efficiency of Food Waste Adsorbents from Synthetic Multi-Element Solutions
Food waste materials are efficient adsorbents for the removal of many metals and metalloids from synthetic multi-element solutions at pH 2.0 and pH 5.5. Table 1
shows that the adsorption capacity varies as a function of the pH and that each food waste adsorbent (200 mg) has a propensity to remove certain elements rather than others. At pH 5.5, As, Ce, Fe, Ga, La, Pb, Sn, Th, Tl, and U precipitate, thus it was not possible to quantify their removal.
At pH 2.0, some elements, such as Mo, Pb, Sb, Sn, Th, Ti and W, were adsorbed in high percentages by most of the food waste materials while others were removed in different percentages depending on the adsorbent exposed. For example, As was removed only by banana peel (adsorption percentages ~37%) and by carob peel, which was able to adsorb more than 75% of the As dissolved in solution. At pH 5.5, the food waste materials resulted less selective for the adsorption of Ag, Cd, Mo, V and Zn, which were removed in high percentages by most of the adsorbents. Decaf coffee waste, coffee waste, potato peel, apple peel, eggplant peel and carob peel resulted more efficient at pH 5.5 than at pH 2.0 for the removal of Ba, Cd, Ni, V and Zn, whereas orange peel appeared more efficient at pH 2.0. W was removed in higher percentages at pH 2.0 by all the adsorbents except watermelon peel. Orange peel and tomato peel removed in higher percentages Ag and Cu at pH 2.0, while coffee waste, decaf coffee waste, apple peel, eggplant peel and carob peel appeared more efficient for the removal of these elements at pH 5.5. Finally, orange peel adsorbed more than 85% of Zn at pH 2.0, whereas apple peel, carob peel, coffee waste and decaf coffee waste were able to remove more than 85% of Zn only at pH 5.5.
Banana peel, watermelon peel and grape waste resulted the most efficient adsorbents for the removal of most of the metals and metalloids from multi-element solutions at pH 2.0 as well as at pH 5.5. Elements’ removal efficiency varies depending on the amount of adsorbent exposed in solution, as expected from the adsorption isotherms defined in previous studies [11
shows that, by increasing the amount of adsorbent (25, 50, 100 and 200 mg) in the multi-element solutions, the adsorption capacity of the food waste materials increases and the selectivity for the elements’ removal decreases. In fact, metals and metalloids compete for certain active sites initially available in lower amounts on the adsorbent surfaces and completely occupied by the elements with the highest affinities. By doubling the amount of adsorbent, the removal percentages gradually increase because the higher availability of functional groups on the adsorbent surfaces progressively reduces the competitiveness between the elements.
In Figure 4
, we observe that the adsorption percentages of some elements (such as Ag, Ce, Mo, La and Pb for grape waste and Ag, La, Mo, U and W for coffee waste) are strictly correlated with the amount of adsorbent exposed. Mo, Sb, Sn and W were removed in high percentages just by 25 mg of grape waste and by 25 mg of coffee waste. These elements favorably competed in the adsorption processes, demonstrating higher affinities to certain functional groups of such food waste materials. Instead Ba, Cd, Co, Ni and Zn were adsorbed in high percentages only by 200 mg of grape waste and watermelon peel which resulted the most efficient food waste adsorbents.
Standard deviations of the results obtained by the adsorption experiments (performed in duplicate) are all under 25%.
Each food waste material showed a propensity to remove certain elements rather than others depending on the functional groups present on its adsorbent surface.
Principal component analysis of the data obtained by the adsorption experiments allowed clustering the food waste materials according to their elements’ adsorption percentages. Figure 5
shows the removal efficiency of the 12 food waste adsorbents, exposed in increasing amounts (25, 50, 100 and 200 mg) to the multi-element solution at pH 2.0.
shows that coffee waste and decaf coffee waste have similar elements’ removal efficiency as well as lemon peel and orange peel which are in the same part of the score plot. Sb, Sn, Mo, W and Ti (on the bottom part of the loading plot) were removed in high percentages by coffee waste and decaf coffee waste (on the bottom part of the score plot). Lemon peel, orange peel, watermelon peel and tomato peel (on the top of the score plot) showed lower adsorption percentages of these elements but appeared more efficient for the removal of the other metals and metalloids (Cr, Th, V, Fe, Cu, Co, In, U, Ba, La, Ni, Ce, Cd, Ag and Pb). Figure 5
shows that increasing the amount of adsorbent exposed, the elements’ removal efficiency considerably increased only for lemon peel, tomato peel, watermelon peel, grape waste and banana peel.
Comparing these results with the results obtained by the multivariate statistical analyses of the FTIR spectra allowed to highlight the potential correlations between the adsorbents’ efficiency and their specific chemical structures. The adsorbents in Figure 5
are marked in the same colors used in Figure 3
to graphically correlate the removal efficiency of the food waste materials with the chemical structures of the adsorbent surfaces. The adsorbents of the red and brown clusters (watermelon peel, tomato peel, lemon peel and orange peel), which showed the highest amounts of −OH of alcohols, phenols and carboxylic acids, of C−H, C−O−C, C−N and P−O of polysaccharides and of C=C of lipids and lignin moieties, resulted more efficient for the removal of Cr, Th, V, Fe, Cu, Co, In, U, Ba, La, Ni, Ce, Cd, Ag and Pb. Instead, the adsorbents of the blue and green clusters (banana peel, apple peel, coffee waste, decaf coffee waste, grape waste, eggplant peel and carob peel), which appeared with lower amounts of these functional groups, showed higher adsorption capacities of As, Zn, Ga, Sb, Sn, Mo, W and Ti.
Therefore from the statistical elaboration of the obtained results we can suppose that the elements above the loading plot’s center of Figure 5
(Cr, Th, V, Fe, Cu, Co, In, U, Ba, La, Ni, Ce, Cd, Ag and Pb), which were removed in higher percentages by watermelon peel, tomato peel, orange peel and lemon peel, have higher affinities to −OH of alcohol groups and C−H, C−O−C, C−N and P−O of polysaccharides than the elements below the loading plot’s center (As, Zn, Ga, Sb, Sn, Mo, W and Ti).