1. Introduction
Lignocellulosic biomass is considered as the most abundant source of renewable carbon on Earth, with an estimated annual production of 2·10
11 tons, mainly represented by residues from forestry and agriculture [
1]. Given its nature as a renewable resource, lignocellulosic biomass is considered as a promising alternative to fossil sources to obtain chemicals [
2,
3,
4] and biofuels [
5,
6]. Through an acid-catalyzed dehydration process, it is possible to convert the carbohydrate component of biomass (cellulose, hemicellulose, and related subunits such as glucose, xylose, galactose, arabinose, etc.) into important chemical intermediates of industrial interest such as furfural, 5-hydroxymethylfurfural (HMF), and levulinic acid (LA) (
Scheme 1) [
2,
7].
However, the exploitation of biomass derived from dedicated crops has led to several problems such as land use changes, the increased cost of raw materials, biodiversity loss, and soil erosion. Therefore, attention has been paid to cellulosic waste and agri-food residues are currently among the most investigated materials, but their exploitation poses the drawback of subtracting these biomass fonts to the extraction of organic and inorganic micronutrients [
8]. Therefore, the discovery of new wastes suitable as cellulose font has become mandatory.
In this context, municipal waste is gaining increasing interest. Among them, cigarette butts, sanitary pants (diapers), newspapers, foods scraps (e.g., peels, grasps), and other residues have gained attention due to their large proportion, which is estimated to be in the order of megatons (Mt) per year all over the world [
9]. In addition, these cellulosic matrices are still unexplored because they are considered as particularly dirty waste (cigarette butts and sanitary diapers), or have other end of life (newspaper is recycled, but in a very polluting processes) or are alternatively burned to produce energy.
The strong Brønsted acids H
2SO
4, HNO
3, and HCl are the mostly efficient catalysts for the hydrothermal deconstruction of cellulose, leading to the top fine chemicals (HMF, LA, etc.), but they impose plant corrosion problems at the industrial level [
10,
11]. Lewis acids such as AlCl
3 are also suitable catalysts for these processes [
10], but present the drawbacks of being irritating to the respiratory system, are corrosive, moisture sensitive, and neurologically harmful [
12].
In contrast, weak Brønsted organic acids such as CH
3COOH or analogous less aggressive Lewis acids like scandium(III) triflate represent preferable alternatives due to the lower toxicity and corrosiveness [
13]. Notably, Sc(OTf)
3 is preferred to analogous lanthanide triflates (e.g., erbium triflate) that are very efficient in the hydrothermal conversion of cellulose, but have been declared as critical materials [
14] by the new rules of EU and green chemistry [
15].
Table 1 reports some representative examples of the catalytic conversion of carbohydrates into valuable chemicals such as furfural and 5-HMF, showing how CH
3COOH and Sc(OTf)
3 are efficient and selective compared to strong acids (entries 1, 4), even if simple monosaccharides are the preferred feedstock, while polysaccharide cellulose commonly requires more aggressive catalysts (
Table 1, entry 2).
Following our ongoing interest in developing green methods obeying circular economy principles [
20,
21], the aim of this work was the development of a protocol that reached two important advantages: (i) exploits municipal waste never used before (e.g., used pants and diapers, newspaper, and soybean peels) as a source of cellulose, and (ii) makes the hydrothermal treatment of cellulose more sustainable by employing less aggressive acid catalysts that are also non-critical materials such as H
3PO
4 [
11], CH
3COOH, and scandium(III) triflate [
19], in order to obtain precious chemicals (HMF, LA, AMF, and furfural) [
2,
22].
2. Materials and Methods
Materials. Ethyl acetate (>99%) was purchased by Honeywell, phosphoric acid (85%), acetic acid (>99.8%), Sc(OTF)3, and all reagents and solvents were purchased from Sigma Aldrich and used without any further treatment.
The four waste cellulosic matrices investigated, namely, cigarette butts (“Rizla + ultra slim 5.7 mm” composed of 98% cellulose acetate [
11]), soybean peels (lignocellulose biomass), newspapers (composed mainly of cellulose), and Fater cellulose (diaper cellulose (composed of cellulose more 70% and 30% super absorbent polymers, personal communication of Fater group SpA) from used sanitary pants gifted by the Fater group Sp A, (Pescara, Italy), were finely chopped into small pieces.
Acid catalysts H3PO4, CH3COOH, and scandium(III) triflate were dissolved into aqueous solutions. CH3COOH was used in two different concentrations of 4 M and 5.7 M, while phosphoric acid (H3PO4) was used at 15% w/w (1.53 M). A total of 50 mg of scandium(III) triflate was dissolved in 15 mL of water and used as a catalyst for all of the substrates in a concentration of 6.7 × 10−3 M.
Instrumentation. GC–MS analyses were run on a Shimadzu GLC 17-A instrument (Shimadzu, MI, Italy) using a SLB-5MS column (30 m × 0.25 mm id, film thickness of 0.25 μm). Mass spectra were performed in EI mode (70 eV) and yields of LA, HMF, AMF, and furfural were determined via GC–MS by means of calibration curves (see
Supplementary Materials). ATR-FTIR spectra (Perkin elemer Waltham, MA, USA) were carried out on a Perkin-Elmer UATR-Two spectrophotometer instrument equipped with a single reflection diamond ATR crystal (refractive index of 2.4). Spectra were acquired with 32 scans in the range 4000–600 cm
−1 by applying both the baseline and the ATR corrections. NMR spectra were recorded on a 500 MHz spectrometer: (Bruker, Milan, Italy)
1H NMR (500 MHz) spectra were referenced to the residual isotopic impurity of CDCl
3 (7.25 ppm) and the
13C-NMR (125 MHz) spectra were referenced to 77.00 ppm. Laser confocal scanning microscopy analyses were performed with an LSM-510 confocal microscope (Zeiss, Milano, Italy).
Typical procedure for hydrothermal treatment. Weighed amounts of cellulose-based waste matrix was suspended into an aqueous solution of the acid catalyst, charged into a 300 mL stainless steel autoclave equipped with a magnetic bar and heated for the proper temperature (in the range 160–200 °C) and time (2–3 h). After cooling, the mixture was filtered and/or centrifugated to separate solid “humins”, which were dried and weighed to give a yield from 20 to 80% (depending on the reaction conditions), while the supernatant was extracted with ethyl acetate (2 × 20 mL).
Purification procedure. The supernatant obtained from the reactions listed in Table 4, entry 1 was extracted with ethyl acetate (2 × 20 mL). Ethyl acetate, containing HMF and AMF, was washed with a 10% solution of sodium bicarbonate (2 × 20 mL) in order to remove the remaining acetic acid. The organic phase was distilled in vacuum to give the blended (HMF and AMF) product. In this case, the product was separated with a column on silica, using hexane/ethyl acetate 2:1 as the mobile phase, and giving Rf = 0.20 for HMF and Rf = 0.58 for AMF.
The reactions listed in Table 4, entries 2 and 5 and Table 5, entry 1 was repeated on the gram scale (2.5 g of substrate) in order to validate the protocol and calculate the e-factors.
Each product was isolated and characterized without further purification and the spectra were in agreement with the literature. The spectra of levulinic acid were previously reported [
11] (see
Supplemental Materials).
AMF (5-Acetoxymethyl-2-furaldehyde): Ref. [
23] colorless oil, GC/MS (70 eV)
m/
z (rel. intensity). 168.30 (M
+, 0.5), 158.30 (0.45), 142.35 (2.94), 127.20 (5.28), 126.15 (100.00), 109.10 (8.04), 97.10 (37.11), 79.05 (26.58), 53.10 (17.79), 45.00 (13.17), 43.05 (64.27). FTIR spectrum (neat) (ν, cm
−1): 3120, 2940, 2834, 1742, 1681, 1582, 1432, 1372, 1275, 1230, 1025, 986, 945.
1H NMR (500 MHz, CDCl
3) δ 9.61 (s, 1H), 7.24–7.12 (m, 1H), 6.57 (d,
J = 3.5 Hz, 1H), 5.10 (s, 2H), 2.09 (s, 3H);
13C NMR (125 MHz, CDCl
3) δ 177.81, 170.32, 155.42, 152.34, 122.48, 112.55, 57.79, 20.89.
HMF (5-(Hydroxymethyl)furfural) [
24] as a dark orange oil, GC/MS (70 eV)
m/
z (rel. intensity): 126.95 (4.56), 125.95 (M
+ 56.59), 108.95 (8.35), 97.95 (5.68), 96.95 (98.97), 68.95 (42.61), 53.00, (20.21), 51.00 (17.75), 43.00, (3.88), 42.00 (8.71), 41.00 (100). FTIR spectrum (neat) (ν, cm
−1): 3122, 2931, 2844, 2718, 1683, 1370, 1280, 1191, 1072, 1023.
1H NMR (500 MHz, CDCl
3) δ: 9.58 (s, 1H), 7.21 (d,
J = 3.5 Hz, 1H), 6.51 (d,
J = 3.5 Hz, 1H), 4.71 (s, 2H), 2.27 (s, 1H);
13C NMR (125 MHz, CDCl
3) δ 177.65, 160.53, 152.36, 122.72, 109.97, 57.63.
Furfural [
25] colorless oil. GC/MS (70 eV)
m/
z (rel. intensity): 40.05 (7.16), 41.00 (2.20), 42.00 (5.45), 51.05 (3.23), 67.00 (10.09), 95.00 (90.09), 96.00 (100.00), 97.00 (5.78); FTIR spectrum (neat) (ν, cm
−1) 3149, 2849, 2811, 1778, 1691, 1674, 1474, 1394, 1246, 1157, 1020,
1H NMR (500 MHz, CDCl3): δ = 9.66 (d,
J = 0.8 Hz, 1H), 7.69 (t,
J = 0.8 Hz, 1H), 7.25 (dd,
J = 3.6, 0.8 Hz, 1H), 6.60 (dd,
J = 3.6, 1.6 Hz, 1H).
13C NMR (125 MHz, CDCl
3): δ 177.93, 152.35, 148.09, 124.22, 112.57.
Calculations and data analysis. Yields in levulinic acid, HMF, AMF, and furfural were calculated based on the weight of the substrate. This yield was calculated with the ratio: product/s (g) obtained after the reaction/substrate(g) × 100. The grams of products were obtained using the GC calibration curves in the
Supplementary Materials.
Regarding the literature and patent reactions, we determined the amount of waste and products (in grams) by using the conversion and molar yields.
Determination of e-factors listed in Table 6 [
26].
2.1. E-Factor of the Reaction Listed in Table 4, Entry 2
Mass of reactants: 30.84 g of CH3COOH (99.8%) in 90 mL of water (the water solvent was excluded from this calculation), cigarette butts 2.5 g; total amount of reactants 30.84 g + 2.45 g = 33.29 g (considering that cigarette butts are composed of 98% of cellulose acetate).
Mass of products: 0.9771 g of 5-AMF + 0.61 g of humins = 0.15871 g
Amount of waste: (33.29 − 1.5871) g = 31.7 g
E-Factor = Amount of waste/Amount of products = 31.7/1.5871 = 19.9
2.2. Determination of E-Factor of the Reaction Listed in Table 4, Entry 5
Mass of reactants: 18.96 g of CH3COOH (99.8%) in 90 mL of water (solvent (water) was excluded from this calculation), soybean peels 2.50 g; total amount of reactants 18.96 g + 2.50 g = 21.46 g.
Mass of products: 0.51 g of Furfural + 1.05 g of humins = 1.56 g
Amount of waste: (21.46 − 1.56) g = 19.9 g
E-Factor = Amount of waste/Amount of products = 19.9/1.56 = 12.76
2.3. Determination of E-Factor of the Reaction Listed in Table 5, Entry 1
Mass of reactants: 0.5 g of Sc(OTf)3 in 90 mL of water (solvent water was excluded from this calculation), cigarette filter 2.50 g; total amount of reactants 0.5 g + 2.45 g = 2.95 g (considering that cigarette butts are composed of 98% of cellulose acetate).
Mass of products: 0.61 g of 5-HMF + 0.7 g of humins = 1.31 g
Amount of waste: (2.95 − 1.31 g) = 1.64 g
E-Factor = Amount of waste/Amount of products = 1.64/1.31 = 1.25
2.4. Determination of E-Factor of the Reaction Listed in Table 6, Entry 7 [14]
Mass of reactants: 12 g of CH3COOH (99.8%) in 60 mL of water (water solvent has been excluded from this calculation), xylose 0.6 g; total amount of reactants 12 g + 0.6 g = 12.6 g.
Mass of products: 0.307 g of furfural
Amount of waste: (12.6 − 0.307) g = 12.293 g
E-Factor = Amount of waste/Amount of products = 12.293/0.307 = 40.04
2.5. Determination of E-Factor of the Reaction Listed Table 6, Entry 4 [15]
Mass of reactants: 0.075 g of cellulose and 0.01 g of AlCl3; total amount of reactants 0.075 g + 0.01 g = 0.085 g.
Mass of products: 0.02325 g of 5-HMF
Amount of waste: (0.085 − 0.02325 g) = 0.06175 g
E-Factor = Amount of waste/Amount of products = 0.06175/0.02325 = 2.65
2.6. Determination of E-Factor of the Reaction Listed in Table 6, Entry 5 [17]
Mass of reactants: 0.040 g of fructose and 0.004 mg of in 2 mL of water (water solvent was excluded from this calculation); total amount of reactants 0.044 g.
Mass of products: 0.01067 g of HMF
Amount of waste: (0.044 − 0.01067 g) = 0.03333 g
E-Factor = Amount of waste/Amount of products = 0.03333/0.01067 = 3.12
2.7. Determination of E-Factor of the Reaction Listed in Table 6, Entry 6 [17]
Mass of reactants: 0.040 g of fructose, 2.0 g of DMSO and 0.004 g of; total amount of reactants 0.040 g + 2.0 g + 0.004 = 2.044 g.
Mass of products: 0.02332 g of HMF
Amount of waste: (2.044 − 0.02332 g) = 2.02068 g
E-Factor = Amount of waste/Amount of products = 2.02068/0.02332 = 86.65