Harvesting of the raw material and preparation of samples

According to the Official Mexican Standard NOM-008-SEMARNAT-1996, which limits the use of the bud of Agave lechuguilla Torr., samples from natural populations were collected during the month of March at the Paredón ejido in the municipality of Ramos Arizpe, Coahuila, Mexico (25°55'47 "N 101°55'47” W), located at an altitude of 728 m, with a mean annual temperature of 33 °C, and a mean annual precipitation of 269 mm; its climate is very dry or semi-warm desert, with a cool winter (^{García, 1973}), and its vegetation type is rosetophilous desert shrub (^{Marroquín et al., 1981}).

The collected samples were washed with tap water in order to remove dirt and other impurities. Once dry, they were cut by hand to a size of approximately 1- 2 cm², and placed in a Koleff-KL10 convection oven until a moisture level of <10 % (p/p) was reached. The dehydrated samples were ground in a Retsch cutting -SM100 mill to a mean particle size of 2 mm. The sample thus obtained was homogenized and stored in plastic containers at room temperature.

Determination of the chemical composition

The extractive determination was performed according to ^{Sluiter et al. (2008}), using ultrapure water and absolute ethanol as solvents in two sequential stages. The cellulose, hemicellulose and lignin were extracted using the procedures of the National Renewable Energy Laboratory (Sluiter et al., 2012), with a modification (^{Mussatto et al., 2011}): 500 mg of the material was hydrolized using H₂SO₄ at 72 % (p/p), during 7 minutes at 50 °C. Subsequently, the hydrolizate was diluted to 4 % (p/p) with distilled water. A second hydrolysis was carried out by autoclaving the reaction mixture at 121 °C for 1 h; the resulting solution was sieved through 0.2-µm filters, and the liquid fraction was analyzed with HPLC in order to determine the levels of glucose (cellulose), arabinose, galactose, mannose and xylose (hemicellulose), according to ^{Mussatto and Roberto (2006}).

The content of acid-soluble (ash-free) lignin was calculated as described by ^{Mussatto and Roberto (2006}). The solid fraction recovered after filtration is considered to be acid-insoluble lignin. The total lignin value was calculated based on the total amount of acid-insoluble and soluble lignin.

The ash content was estimated through incineration of the samples at 550 °C for 4 h (^{Horwitz and Latimer, 2005}).

Antioxidant potential

The extracts (liquors) of A. lechuguilla were obtained using 2 grams (on a dry basis) of the sample with 40 mL of each solvent: 1) methanol, 2) ethanol and 3) acetone at 60 % (v/v) in all three cases. Each treatment was carried out at a temperature range of 60 to 65 °C, during 90 min with magnetic agitation. The control consisted in the same amount of sample subjected to treatment in an autoclave at 120 °C for 90 min, with 40 mL of NaOH at 2 %. The four extracts were then separated by centrifugation at 2 500 rpm for 20 min and sieved through 0.22 μm filters.

The antioxidant capacity of the extracts was determined by two spectrophotometric methods: the stable free radical assay (DPPH) and the ferric reducing antioxidant power assay (FRAP) were adapted for a 96-well microplate (^{Martins et al., 2012}). The total antioxidant capacity (TAC) was evaluated using the method described by ^{Prieto et al. (1999}).

Structural characterization

The FTIR spectra were recorded with the FTIR spectrometer (Agilent, Cary 630, USA); the FTIR in the band mode of absorption were in the range of 650-4 000 cm^-1, with a resolution of 4 cm^-1 and 32 explorations. Consecutively, the resulting FTIR spectra were analyzed and standardized and are represented in the software Origin (^{OriginLab®, 2016})

Surface morphology

A. lechuguilla particles were observed using scanning electron microscopy (SEM), in a microscope with a field-emission electron gun at an ultra-high resolution (Nova 200 Nano SEM, FEI Company). Prior to the analysis, the samples were covered with a thin film (35 nm) of Au-Pd (80-20 % by weight). The images were obtained by applying a voltage of 10 kV acceleration, with increases of 200 and 2 000.

Analysis

All the experiments were made in triplicate and the recorded values represent the mean for each treatment, with its corresponding standard deviation.

Chemical Groups and Bonding Layout of the Components

The FTIR analysis confirmed that the buds of A. lechuguilla have absorption bands that are typical of lignocellulosic materials, although their magnitudes differ (^{Adel et al., 2010}).

Because of their complex nature, their FTIR spectra are often separated into two main sections; -OH and -CH, which extend through the region of 3 800-2 700 cm^-1, and the digital fingerprint region that is assigned to the stretching vibrations of various groups of components of these lignocellulosic materials to 1 800-800 cm^-1 (^{Poletto et al., 2012}).

The broad peak between 3 400 cm^-1 is related to the O-H stretching vibrations of the hydroxyl group. The signals between 2 920 cm^-1 and 2 850 cm^-1 are commonly observed in samples of lignocellulosic biomass and correspond to the methyl and methylene groups (-CH₃ and CH₄), respectively. However, for A. lechuguilla samples these signs appeared at 2 914 cm^-1, which may be due to the greater extractive content of this raw material, as some compounds in organic extractives, such as methyl esters of fatty acids and esters of phenolic acid, contain methyl and methylene groups (^{Poletto et al., 2012}).

The signals recorded at 1 735 cm^-1, 1 375 cm^-1, 1 240 cm^-1, 1 165 cm^-1 and 1 060 cm^-1 are characteristic of the stretching and twisting of various groups of carbohydrates. For the A. lechuguilla bands attributable to symmetric and asymmetric stretching of the C-O-C in glycosidic bonds, they occurred at 1 100 cm^-1 and 1 095 cm^-1 and correspond to the β-1,4-glycosidic bonds that unite the glucose units in cellulose and hemicellulose (Figure 1).

Figure 1 FTIR spectrum of the bud of Agave lechuguilla Torr. 

Surface morphology

The images obtained using scanning electron microscopy revealed a rigid structure in A. lechuguilla (Figure 2). The small fractures observed can be attributed to the milling process (Figure 2a), and the small crystalline forms (Figure 2b) would be small crystals of calcium oxalate, commonly present in this type of biomass (^{Pérez-Pimienta et al., 2016}).

a) Fractures of the fibers; (b) crystalline forms present in the material.

Figure 2 Electron micrographs by scanning electron microscopy (SEM) of crushed leaves of Agave lechuguilla Torr. 

Chemical composition

After the elimination of the extractive compounds, the material lost 45.34 % of its original weight, which corresponds to the non-structural components such as sucrose, nitrate/nitrites, protein, ash, chlorophyll, and fatty compounds of the plant (^{Sluiter et al., 2012}).

The relative moisture of the material, after the drying process, was 4.33 %. The cellulose fraction was calculated based on the glucose detected using HPLC (0.55 g L^-1). The hemicellulose was determined on the basis of the concentration of xylose (0.22 g L^-1), galactose (0.18 g L^-1), arabinose (0.068 g L^-1) and mannose (0.08 g L^-1). The cellulose, hemicellulose, and lignin contents (dry basis) in the samples were 17.72 %, 17.15 and 7.32 %, respectively (Table 1).

Antioxidant potential

The antioxidant potential of A. lechuguilla was assessed using three of the most common testing methods (DPPH, FRAP and CAT). DPPH has been widely used as a fast, reliable and reproducible parameter to measure the in vitro antioxidant capacity of pure compounds as well as extracts from plants (^{Burda and Oleszek, 2001}; ^{Ara and Nur, 2009}). The results of the assay DPPH are shown in Figure 3; the values are provided in terms of % of inhibition by DPPH. It is evident that the treatments with solutions of methanol, ethanol and acetone were of the order of 89.47 %, 90.09 % and 89.75 %, respectively; while the liquor from the treatment with NaOH had a low DPPH value (32.148 %) (Figure 3).

Figure 3 Percentage of inhibition by DPPH reagent of different extracts of Agave lechuguilla Torr. 

The results expressed in mM Fe(II) g^-1 biomass (on a dry basis) were 0.51, 0.51, 1.6 and 0.46, respectively, for the extracts obtained through treatment with methanol, ethanol, acetone and NaOH respectively.

Although the solutions used in the treatments are binary water-solvent mixtures, not all have the same physical-chemical characteristics, since the polarity of each solvent (or mixture of these) has a direct relationship to the amount and quality of antioxidant compounds that can be extracted by it (^{Aliyu et al., 2012}). The order of hierarchy in terms of polarity for this work is Acetone> Methanol>Ethanol. This is related to the concentration of antioxidants detected using the FRAP method in liquors resulting from the extractions with solvents, as the highest concentration was obtained with the acetone method (1.60 mM Fe(II) g^-1 of biomass), and the values of the liquors obtained with methanol and ethanol, both of which exhibit a very similar polarity, were 0.51 mM Fe(II) g^-1 of biomass. Finally, although no solvent was used to extract it, the biomass treated with NaOH had a value of 0.46 mM Fe(II) g^-1 because a fraction of antioxidant compounds were extracted by virtue of the interaction of the temperature used and the water present in the solution.

Figure 4 Antioxidant capacity of various extracts of Agave lechuguilla Torr. detected using the FRAP method. 

Figure 5 shows the results for the total antioxidant capacity measured with the phosphomolybdenum method, which is based on the reduction of Mo (IV) to Mo (V) for the extract, with the subsequent formation of the green phosphate Mo (V) complex at an acidic pH. This model evaluates the antioxidant capacity of both water-soluble and fat-soluble compounds, hence the name “total antioxidant capacity” (^{Aliyu et al., 2012}). The records indicate that there is a relationship between the method of extraction and the solvent used in the concentration of antioxidants in the resulting liquor; that is to say, that the extraction using the acetone method concentrated 1.75 mg of antioxidant compounds equivalent to α-tocopherol, which are capable of reducing the oxidant compound in the reaction matrix.

Figure 5 Antioxidant capacity of various extracts of Agave lechuguilla Torr. detected using the Total Antioxidant method. 

Each of the applied methods evaluates the antioxidant potential of a biological sample in different ways, and each one of these is based on a variety of reactions. Despite the fact that the DPPH assay showed little difference in the three different extractions, there is a correlation between the results of the total antioxidant capacity and testing with FRAP, in which the acetone extract exhibited higher values than the methanol and ethanol extracts. This would mean that, under the same conditions, the antioxidant compounds detected in A. lechuguilla have more affinity with acetone. Furthermore, the temperature conditions of the extractions were favorable for the acetone, whose boiling point is lower than that of methanol and ethanol; thus, higher vapor pressures are generated in the containers, which leads to higher gas-solid extraction rates.

Lower antioxidant capacity values were obtained for the sample treated with NaOH than for all the other samples. The difference between the samples treated with solvents and the sample treated with NaOH may be attributed to the longer operation time allotted to the former, as, according to the literature, the extraction of this type of compounds depend not only on the temperature but also on the time during which the herbaceous samples are in contact with the solvent (^{Ben Hamissa et al., 2012}). While for the treatment with NaOH, the temperature was higher and the total operation time of the process was longer than for the treatments with solvents, the reactions involving antioxidants extracted from plants are pH dependent, as electron transport is limited in a highly oxidized environment -i.e. in an environment saturated with OH-, which tends to degrade phenolic compounds and to generate highly oxidizing substances due to delignification; therefore, the resulting liquor will have a lower antioxidant capacity (^{Yokoyama et al., 2007}; ^{Aliakbarian et al., 2009}).